Textual Voices in Corporate Reporting: A Cross-Cultural Analysis of Chinese,Italian,and American CSR Reports

Abstract

This article investigates direct quotations in a corpus of corporate social responsibility (CSR) reports in Italian, Chinese, and English. The corpus is composed of 60 CSR reports published by Italian, Chinese, and American companies in the banking and energy sector. The study aims at exploring what types of textual voices are involved in the discourse of CSR reporting and how different sources of voices are represented, using the framework of social actor representation proposed by Van Leeuwen. The results show that the voices presented in direct quotations are often “orchestrated” by companies into “symphony” rather than “polyphony.” Most of the sources of direct quotations are represented as individuals with specified names. The comparative analysis shows that companies from different cultural backgrounds present different preferences in selecting and representing the various sources. The Italian and American CSR reports present more voices from managers, while the Chinese CSR reports show a clearer preference for voices from employees and clients.

Keywords

direct quotations CSR reporting Italian Chinese English

Introduction

People’s lives have been deeply affected by the negative impacts of corporations on society, for example, by the crisis in the financial sector or the environmentally destructive activities of energy companies. One of the consequences of this has been a crisis of trust between corporations and their stakeholders. The term stakeholder is used here in the sense defined by Freeman (1984), who pioneered stakeholder studies in management and defined stakeholders as “any group or individual who can affect or is affected by the achievement of the organization’s objectives” (Freeman, 1984, p. 46). In external communication, businesses need to manage increasingly complex relationships with various stakeholder groups and to integrate their different voices. This need is witnessed by academic and social debates surrounding the emerging of corporate social responsibility (CSR) issues and CSR communication genres.

The present study investigates the genre of the CSR report, with a focus on how different voices are represented by companies, and looks in particular at the presence of direct quotations in the report. Forms of speech and thought representation in texts have long been of interest to many different areas of language studies, ranging from stylistics to pragmatics and cognitive studies (see, e.g., Bublitz & Bednarek, 2006; Leech & Short, 1981/2007; Rosier, 2008; Semino & Short, 2004; Tannen, 1989/2007; Toolan, 2006; Vandelanotte, 2009). The use of quotations has been studied in conversational speech (e.g., Blackwell, Perlman, & Tree, 2015; Clark & Gerrig, 1990; Günthner, 1999; Tannen, 1989/2007; Tree & Tomlinson, 2007), in academic discourse (e.g., Bondi & Silver, 2004; Fløttum, Dahl, & Kinn, 2006; Hyland, 1999; Petrić, 2007; Swales, 1986; Thompson & Tribble, 2001; Thompson & Ye, 1991), in the press (e.g., Calsamiglia & Ferrero, 2003; Sleurs, Jacobs, & Van Waes, 2003; Smirnova, 2009; Yung, 1996), and in political speech (e.g., Bevitori, 2005; Gruber, 2015; Kuo, 2001; Shukrun-Nagar, 2009), while we have not been able to find any study focusing on the genre of corporate reports.

This study aims to explore the use of direct quotations in the CSR report. The corpus is composed of 60 CSR reports published by banks and energy companies from Italy, China, and the United States. We aim to address the following research questions:

Research Question 1: What types of textual voices are involved in the CSR report?

Research Question 2: How are different sources of voices represented in the CSR report?

Research Question 3: What insights does the cross-cultural comparison provide on CSR discourse?

The rest of the article is organized as follows. The second section sheds light on the polyphonic aspects of CSR communication. The third section focuses on textual voices and provides an account of how sources can be represented according to Van Leeuwen’s framework on social actor representation (Van Leeuwen, 2008). The fourth section describes the method and the corpus, drawing on techniques developed in grounded theory. The fifth section analyzes the sources of direct quotations found in our corpus in terms of their relation to the company and of their social or professional role. The sixth section investigates how voices are used and represented in CSR reports, and the seventh section presents cross-cultural similarities and variations in the use of direct quotations. In the final section, we provide a discussion and answers to the research questions.

Polyphony in CSR

Corporate social responsibility refers to “the responsibility of enterprises for their impact on society,” as put forward by the European Commission in the renewed EU strategy 2011-14 for Corporate Social Responsibility (European Commission, 2011, p. 6). To meet their social responsibility, “enterprises should have in place a process to integrate social, environmental, ethical, human rights and consumer concerns into their business operations and core strategy in close collaboration with their stakeholders” (European Commission, 2011, p. 6). This multiplicity of stakeholder issues certainly influences the polyphonic nature of CSR communication.

Christensen, Morsing, and Thyssen (2011) define the notion of polyphony in CSR as an idea that combines several voices into a complex concept, opening a space for individual voices to be independent, instead of being integrated into or dominated by other voice(s) in forms that Castelló, Morsing, and Schultz (2013, p. 688) would call monophonic or homophonic. In this context, polyphony refers to the diversity and differences (and conflict) of voices that interact with one another dynamically, with a focus on corporations’ management and integration of diversity, incompatible values, and dissent at a discursive level (Castelló et al., 2013, p. 688).

This conflicting CSR interaction and “dynamization” is made particularly visible in networks of media such as blogs and fora (Castelló et al., 2013, p. 688), but it seems unlikely to be equally visible in traditional corporate-centric communication such as press releases and corporate reports. In our corpus of CSR reports, in fact, although the companies do give voice to some stakeholders creating textual “polyphony,” the voices presented are always aligned with the companies, and never challenge their performance or idealistic claims. From this point of view, the companies in our corpus could be said to be orchestrating a “symphony,” rather than a “polyphony.” The notions of “symphony” and “polyphony” refer here to different aspects of the argumentative structure of the text. As defined by van Eemeren and Grootendorst, “[a]rgumentation is a verbal, social and rational activity aimed at convincing a reasonable critic of the acceptability of a standpoint by putting forward a constellation of propositions justifying or refuting the proposition expressed in the standpoint” (van Eemeren & Grootendorst, 2004, p. 1). The notion of symphony, rather than polyphony, can be used to refer to argumentation that does not have a strongly dialectical structure, that is, a structure including both arguments and counter arguments. In other words, symphony differs from polyphony in that it does not present propositions “refuting the proposition expressed in the standpoint” (van Eemeren & Grootendorst 2004, p. 1), but rather, it makes its point by supportive argumentation only.

It is our contention that this presence of “different” voices, although it cannot be defined as “monophonic,” is essentially symphonic: The different voices share the same stance and are all aligned with the company.

Toward a Methodological Framework for the Analysis of Quotations in CSR Reports

Voices in CSR Reports

The discussion of textual voices is based on a preliminary distinction between averral and attribution (Sinclair, 1988). Averral refers to propositions put forward by the writer, while attribution denotes the case where a proposition is indicated as deriving from a source. In the case of CSR reports, the text that is directly averred (not attributed to any other source) represents the voice of the company, whereas the text that is in various ways attributed to others represents other textual voices.

In traditional definitions, in terms of message types, attributions can be realized by direct quotations, without using any paraphrases, or by recourse to paraphrasing strategies in different forms of indirect quotations. The present study focuses on the use of direct quotations, that is, attributed voices that are quoted directly. This focus on direct quotations has two reasons. First, investigating the CSR reports in our corpus, we noticed that the use of direct quotations is remarkable, while indirect quotations were rarely noticed. Second, for time limitations and practical considerations, we restricted the research aim to investigating only direct quotations, which are often presented with quotation marks or other visual strategies, instead of indirect quotations, which do not have noticeable formal or visual hints. In CSR reports, direct quotations are presented not only with quotation marks (inside or outside the running text) but also, preferably in text segments, with different font type, which distinguish them from the main text, or in an independent box. This reduces to a minimum the use of reporting verbs or nouns, which would often guide the reader to an interpretation of the nature and function of the attributed text (e.g., Thompson & Ye, 1991). The mere juxtaposition of the quoted text and the main text makes the relationship between the two mostly inexplicit.

From the point of view of textual rhetoric, quotations are like mini-genres embedded in a genre. They can convey and perform different “moves” (Bhatia, 1993; Swales, 1990), that is, rhetorical units with specific communicative purposes or functions to serve the overall communicative purpose of a genre. For example, a direct quotation can be used to provide information, to establish credentials, or to describe external circumstances, as part of different moves.

Unlike genres addressed to one person (e.g., personal letters or messages) or one community (e.g., textbooks, guidelines), the CSR report is characterized by a participation framework (Goffman, 1981) with a wide range of addressees, including individuals and organizations that may be affected by the consequences of a company’s activities. In dealing with this complexity of addressees, the presence of voice multiplicity may contribute to enhancing the credibility and creating consensus among the readers.

The leading voice of the CSR report is the voice of the company, which has the aim to “express” its CSR commitment, strategy, and practice. Other voices include those of internal stakeholders, such as managers, executives, employees, shareholders, and other owners (internal voices), and those of external stakeholders (e.g., customers, communities, suppliers, partners, creditors, government), external evaluators, auditors, and so on (external voices).

The internal voice of senior decision maker(s) of a company is often presented in a signed “letter to stakeholders” or in the form of Q&A or interview. The CEO’s letter is frequently employed in corporate reporting genres such as CSR reports as well as annual reports. As Garzone (2004, p. 322) observes, the CEO’s letter in annual reports has the communicative function to “provide a concise account of a justification for the company’s activities and performance.” The presence of the CEO’s voice contributes to the authoritativeness of the justification discourse, thereby improving and enhancing the credibility of the corporate report.

The auditor’s report is an independent and external opinion and review intended to certify the CSR information disclosed by the company. It represents a clearly sourced external voice aiming to enhance the credibility of the report content.

The information carried by the two voices is formulated in part genres—the decision maker(s)’s statement (internal voice) and the auditor’s report (external voice) that are somewhat prescribed in the GRI (Global Reporting Initiative) guidelines, the most common guidelines for CSR reporting. It should be noticed that the CEO’s voice and the auditor’s voice are actually necessary elements, that is, felicity conditions of the report as complex speech act. A CSR report in which the managers who direct the company do not sign explicitly, endorsing the report’s content, or in which no auditor’s report is presented, may not be regarded as legitimate. This nature makes these two voices different from other textual voices that will be discussed further. Apart from these two prescribed and obligatory voices, CSR reports sometimes present voluntary use of other textual voices, with either internal or external source, often textualized in direct quotations.

In defining our research question, we noticed that the rhetorical strategy to present voices from various sources seems to characterize the CSR report vis-à-vis other corporate genres such as the annual financial report. As an illustration, the 2013 annual report of ABC (Agricultural Bank of China) does not use any direct quotation, while its 2013 CSR report presents 44 instances. The analysis that follows will look at how direct quotations are attributed and represented in the CSR report.

Quoted Sources

When looking at the sources quoted, it is important to focus on who they are and how they are represented. Following Thompson (1996), textual voices can be classified as specified other(s), unspecified other(s) (e.g., one of the women said . . .), or the community (e.g., proverbs), and so on. (Thompson, 1996, 506-511). The present study focuses on the representation of specified or unspecified sources.

Types of quoted source (e.g., Smirnova, 2009), or speaker (e.g., Bondi & Silver, 2004), agent (e.g., Calsamiglia & Ferrero, 2003), quotee (e.g., Kuo, 2001), referents (Clark & Gerrig, 1990) can attribute different rhetorical and argumentative force to the quotations. Smirnova’s (2009) study of quoted sources in U.K. newspaper discourse shows that professional characteristics (e.g., position, profession, academic degree) seem to possess a much stronger argumentative force than personal characteristics (e.g., nationality, age). The use of different types of quoted source can also reflect the importance of the role of the quoted source as perceived by the reporter or by the community where the genre is located. Calsamiglia and Ferrero (2003), for example, suggest that scientific voices had much less weight than political ones in the press coverage of the “mad cows” case.

In the CSR report, the types of quoted sources can be categorized into internal sources (e.g., the CEO, managers, employees), external sources (e.g., customers, authorities), and unspecified sources. Our hypothesis is that different cultures may be characterized by specific preferences for the quoted source.

In addition to these major types of quoted source, the present study also looks at how sources are represented by companies, using a critical discourse analysis approach. Van Leeuwen (2008) has presented a framework of the representation of social actors in discourse (see Appendix 1 for a summary of the various categories of social actor representation that he has proposed; all appendixes are available in the online version of the article). The sources of quotations in our corpus have been studied considering the representation type of Category 7 and Category 10 in Van Leeuwen’s (2008) framework (Table 1).

Table 1.

Types of Social Actor Representation Considered.

Category 7	1. Genericization 2. Specification 2.1 Individualization 2.2 Assimilation 2.2.1 Collectivation 2.2.2 Aggregation
Category 10	3. Indetermination 4. Determination 4.1 Categorization 4.2 Nomination

The categories were selected because they appeared to be the most suitable to reveal how companies present different quoted source, as the other representation types such as “passivation,” “abstraction,” and objectivation” did not seem relevant to the representation of quoted source.

On the level of Category 7, social actors can be categorized as “classes” ( generification , e.g., Cats are cute), or as specific, identifiable entities ( specification , e.g., My cat is here) (Van Leeuwen, 2008, p. 35). Specific entities can be presented either as individuals (individualization, e.g., My cat is here) or as “groups” (assimilation, e.g., The cats in my hometown are smart) (Van Leeuwen, 2008, p. 37). A group is different from a class: A class is the set of all individuals of a certain kind (“Cats” meaning “All cats, cats in general”), whereas a group is a subset of individuals (“the cats in my hometown”). When social actors are represented as groups, they can be treated as “statistics” (aggregation, e.g., 45% of the cats are female) or simply “collectivized” (collectivation, e.g., The cats in my hometown are smart) (Van Leeuwen, 2008, pp. 37-38).

On the level of Category 10, social actors, generalized or specified, can be represented as unspecified, “anonymous” individuals or groups (indetermination, e.g., Someone is there), or they can be represented with specified identify (determination, e.g., My cat is here). For determination, social actors can be represented by being “nominated” (nomination, e.g., my cat Sisi ) or “in terms of identities and functions they share with others” (categorization, e.g., my ca t).

Corpus and Procedures

Corpus sampling procedures were combined with methods developed in grounded theory (Glaser & Strauss, 1967; Strauss & Corbin, 1994, 1998), “a general methodology for developing theory that is grounded in data systematically gathered and analyzed” (Strauss & Corbin, 1994, p. 273). We used this methodology for description and “conceptual ordering” (organizing of data according to a set of properties; Strauss & Corbin, 1998, p. 15), providing a set of related concepts and assumptions as a basis for theory development. Two essential operations in this kind of analysis are “asking questions” and “making comparisons” (Strauss & Corbin, 1998, p. 73), which are analytic tools used in the process of microanalysis of samples. In our study, the microanalysis of data is combined with “open coding,” which refers to “[t]he analytic process through which concepts are identified and their properties and dimensions are discovered in data” (Strauss & Corbin, 1998, p. 101). Moreover, the process of data gathering and coding was guided by the principle of “theoretical sampling” (Strauss & Corbin, 1998, p. 201). In other words, the data gathering was driven by and integrated with concepts and patterns emerging from the process of microscopic examination of data.

In general, the procedures used in gathering and analyzing data are as follows. These will be described with more details in the following pages.

Building a corpus on the basis of practical considerations

Analyzing and coding data to generate concepts for theoretical sampling

Making comparisons and assumptions for theory evolving

Corpus Building

The corpus (CSR-ICU, I for Italy, C for China, U for USA) is composed of 60 CSR reports of 2013 published by major banks and energy (oil, gas, electricity) companies in Italy, China, and the United States. The reports under investigation are in the original language: 20 reports in Italian, 20 in Chinese, and 20 in English. The corpus includes the top 10 banks and top 10 energy companies in each country, whose report was available for download on a set date, according to ranking lists provided by www.relbanks.com and www.top250.platts.com.¹ A list of the companies can be found in Appendix 2.

Initially, the research aimed at examining potential rhetorical similarities and variations in the use of direct quotations in Italian, Chinese, and English. With a focus on the three languages, a pre-study analyzed a small corpus with six CSR reports in Italian produced by Italian companies, six in Chinese produced by Chinese companies, and six in English produced by companies based in the United Kingdom or in the United States. However, after a close examination of the pre-sample, we based our choice on cultural background rather than linguistic conventions. Although the CSR report is mostly a voluntary tool for corporate CSR communication, it is encouraged and promoted by governmental organizations. Therefore, this genre may be highly affected by the national legal and cultural framework where the headquarters of a company is based. With this consideration, for the English sample we decided to narrow the scope and focus on one country only, that is, the United States, on the assumption that it would present more interesting cultural variations when compared with Italy as a European country. This consideration has also guided the sampling of Chinese CSR reports: We excluded companies with headquarters in Hong Kong. This is a particular case because its business culture is deeply influenced by the Western world and might not be the best choice to represent Chinese business culture in general.

The banking and energy sectors were chosen for the noticeable number of companies that publish CSR reports. The corpus could be larger if we included more sectors. We did not do this because in the pre-study, we noticed that, CSR reports might be influenced not only by the national culture but also by a specific sector culture (which is evidenced in the present study). Therefore, we focused on two sectors only to reduce variables that might influence the results.

Data Coding

The coding process — supported by Notetab Light (http://www.notetab.ch/ntl.htm) — focuses on three dimensions: the identification of direct quotations, the categorization of source types, and the annotation of source representation types.

The sampling of direct quotations is a dynamic process, which at the same time contributed to the delineation of properties of direct quotations. In the previous section, we defined direct quotations as attributed voices directly quoted with specified or unspecified source. This definition is concise and seems clear enough, but it is not sufficient when applied to the complexity of real data. There were at least two cases when the decision of what to include in the study was specifically determined by our research questions.

The first one refers to the CEO’s letter and the auditor’s report. Unlike simple and short genres such as an invitation letter, the CSR report is a complex and extensive genre that often encompasses multiple supplementary part-genres, among which are the CEO’s letter and the auditor’s report (see section “Voices from Stakeholders and Non-Stakeholders”). Although these two texts are represented as attributed to specific voices, they were not considered in our sample, because they have become prescribed and regular documents in the CSR report, and are not so significant to our research questions, which focus on the rhetorical use of direct quotations in the CSR report.

Another ambiguous case is the use of “pseudo quotations,” which refer to words obviously proposed by the company but depicted as voices from stakeholders (see Figure 1 below). These are presented in the form of direct quotations, which do not have a specified source. Although in this case the direct quotation becomes a purely rhetorical choice, our sample considered this phenomenon, because it is interesting for a study on how companies present, represent, or establish various voices in the rhetorical construction of the CSR report. In total, 472 instances of direct quotations were found. The basic information of the corpus is presented in Table 2.

Figure 1.

An extract from A2A 2013.

Table 2.

Basic Information of CSR-ICU.

	Italy	China	USA	CSR-ICU
Tokens (running words)	1107122	471050	499889	2078061
Frequency of direct quotations	66	224	182	472

Note. CSR-ICU = corporate social responsibility, I for Italy, C for China, U for USA.

When the coding of direct quotations was completed, we focused on the categorization of source types and source representation. The focus on source types aimed at exploring the rhetorical importance that companies attribute to different sources. The initial categories included “external stakeholders” and “internal stakeholders,” because at the beginning we attempted to investigate how voices from stakeholders are represented in this genre used for stakeholder communication. However, with the microanalysis of data, we noticed that some sources are neither external nor internal (explicit) stakeholders of the company, such as external researchers and journalists. Moreover, some voices from stakeholders do not address their relationship at stake with companies. Considering these phenomena emerging from the data, we coded the source types with two major categories that emphasize the property of relationship type with the company: internal sources and external sources. Moreover, we further annotated the data with subcategories, which emphasize the property of professional or social role of the sources, such as “managers,” “employees,” “clients,” and so on. This method of categorization is more in line with the real data, and it can contribute to exploring how source types with different relationship and social roles are conceived by companies in the rhetoric of CSR communication.

Furthermore, on the basis of the Social Actor Network proposed by Van Leeuwen (2008, p. 52), we studied how different sources are represented in the CSR report. With a pre-analysis of the data, we decided to focus on the representation types of Category 7 and Category 10 of Van Leeuwen’s framework (see section “Quoted Sources”), because the use of other categories is not able to reveal the significant properties of quoted sources. For example, Category 3 includes the representation types of activation and passivation (Van Leeuwen, 2008, p. 52), but in our case, quoted sources are never represented in passive role; therefore, they were not included in the coding.

Making Comparisons and Assumptions

At this stage, we examined the coded data across the three languages and highlighted cross-cultural characteristics in terms of the use of direct quotations, source types, and the representation of source. Based on the cross-cultural similarities and differences, we derived assumptions on the use of direct quotations, which can be found in the final section.

Voices From Stakeholders and Non-Stakeholders

As mentioned above in presenting the methodological framework for the analysis of quotations, the sources of the direct quotations found in the CSR reports can be divided into three categories: internal source (e.g., managers, employees), external source (e.g., clients, public authorities, local community, evaluator, researchers, partners), and unspecified source (with no explicit attribution).

An important further distinction here is between stakeholders of the company and non-stakeholders. Internal sources are obviously stakeholders (e.g., managers, employees), while external sources and unspecified sources can be either stakeholders of the company (clients, local community, partners) or non-stakeholders (e.g., external researchers, journalists). The distinction proves to be an important one from the point of view of the general discourse functions of CSR quotes.

The voices of stakeholders (whether internal or external) are often used to address and evaluate the commitment of the company to CSR principles and practices. Evaluative language use does not only play a key role in CSR reports in general (Fuoli, 2012); it also seems to characterize many of the quotations included in the reports, expressing the opinion of textual voices and variously appraising the company’s CSR commitment or performance. External stakeholders often do that by (positively) appraising the relationship they have with the company (Example 1).

(1) “AMEC black & McDonald Limited has been working with ExxonMobil Canada for almost 10 years as a contractor. Our cultural growth in safety performance has been powerfully impacted by our relationship with ExxonMobil in an extremely direct and positive way. Our world-class safety record would have not been possible without ExxonMobil’s guidance and support.” (Mark Healy, President, AMEC Black & McDonald Limited (ExxonMobil 2013)

In Example 1, Mark Healy speaks on behalf of AMEC Black & McDonald, which has a business relationship with ExxonMobil. In his words, the use of exclusive first-person pronouns (our) frames his company as a counterpart to ExxonMobil. He evaluates his company’s relationship with ExxonMobil (powerfully impacted by, extremely direct and positive way), and he appreciates ExxonMobil’s contribution to their community (guidance and support).

On the other hand, internal stakeholders (managers in particular), tend to appraise the company’s commitment to CSR directly, without explicitly dealing with their relationship “at stake with” the company (Example 2):

(2) “Citi is dedicated to being a top company for diversity and inclusion everywhere we operate. While we recognize local customs and cultures, we will uphold our commitment to equality and fairness and to provide opportunities for all of our employees to make progress toward personal and professional goals.” (Ana Duarte McCarthy, Chief Diversity Officer, Citi (Citi 2013)

In Example 2, the Officer McCarthy highlights the company’s idealistic and commitment claims speaking in the name of Citi, rather than dealing with the company as a counterpart. This is evidenced by her extensive use of inclusive first pronouns (we and our), representing McCarthy and the company as a unit. The use of our employees (instead of we employees) is a clear marker of identification with the company.

One common feature of Example 1 and Example 2 is that they are both specific instances of an endorsement of the company’s commitment to CSR, with a view to one or more aspects of CSR (e.g., equality and fairness of treatment). Both examples use explicit appraisal to testify the company’s positive performance and thus contribute to enhancing the company’s credibility. In some other cases, the appraising mechanism can be less explicit. The following quotation comes from a 12-year-old girl who has won a painting competition launched by the company.

(3) (Translation by CNOOC) “I wish that one day as the CNOOC workers, I can stand on the high drilling platforms, appreciating the free flying of seagulls, hearing the merry songs of the sea.” (Winner of “My CNOOC Dream” Painting & Calligraphy Competition Wang Xinzhu (12 years old) (CNOOC 2013) (see Appendix 3 for original version in Chinese)

In Example 3, the appraisal is more indirect. The quotation appears in the section of “My CNOOC Dream,” which focuses on stating the company’s values and beliefs. It presents a girl’s aspiration to become a CNOOC worker, and depicts the girl’s positive imagination of working as a CNOOC worker (high drilling platforms, appreciating the free flying of seagulls, hearing the merry songs of the sea), representing the job as a great opportunity and a dream job. The positive description of the working environment (free flying, merry songs) also implies that CNOOC operations are harmonious with the natural environment. By using direct quotation, the text involves the reader in the aspiration and evokes empathy and positive imagination. Even more vividly and emotionally, the quoted text is also accompanied by a “quoted” painting, the one with which the girl has won the competition (see Appendix 4). Even if not explicitly referring to the company, the use of evaluative and positive words in association with the working place in Example 3 acts once again as an endorsement of the company’s commitment to creating positive working environments.

A few direct quotations are attributed to non-stakeholders, which include external evaluators, external researchers/experts, celebrities, authorities, news agencies, and so on. These voices do not address an existing relation with the company. They mainly appraise the company (and the company’s performance) or provide a justification/motivation for the company’s activities. The endorsement of the company on the part of an external party with no interest at stake is obviously an important asset in terms of credibility.

External voices in our corpus also present specific preferences in terms of pragmatic functions. Voices cited from external evaluators, external authorities, and journalists are often quoted when presenting a direct appraisal of the company. Examples 4 and 5 illustrate this:

(4) (Translation by CNOOC) “When I walked into Putian gas-fired power plant, I saw four HRSGs (heat recovery steam generator) lining up neatly without any discharge of smoke dust, which is different from former power plant areas where multiple chimneys stand.” (Journalist of Xinhuanet Li You (CNOOC 2013, Section of “Listening to”) (see Appendix 3 for original version in Chinese)

(5) In remarks about the long-term unemployed, President Barack Obama said, “[T]hat’s why we’re excited to have programs like. . . . Platform 2 Employment . . . [I]t’s great to have these intermediaries and nonprofits who are also able to show success, even with folks who have been out of work for a long, long time.” (Citi 2013, Section of “Financial Security and Enterprise Development”)

In Example 4, the journalist appraises the company directly by showing what he has witnessed (lining up neatly without any discharge of smoke dust) so as to testify the company’s good performance. By citing the journalist’s quotation, the text establishes an interpersonal involvement (Kuo, 2001, p. 181), and provides a sense of immediacy (Tannen, 1989/2007), by shifting the readers’ visual point on the position of the writer to be able to experience directly the voice of the journalist. In Example 5, the bank quotes former President Barack Obama praising a program of the bank (e.g., great, success). The obvious authoritativeness of the source clearly enhances the credibility of the evaluative words.

While some external voices (e.g., news agency, external evaluators, or external authorities) are presented to provide proof for the company’s good CSR performance, voices from other sources (e.g., researchers/experts) tend to present contextual information that provides justification for the company’s action. An analysis of the pragmatic and textual functions of quotations in CSR reports lies beyond the scope of the present article, but it might be noticed at least that quotations—very frequently not integrated in the running text but presented in boxes—may contribute in different ways to the report, by witnessing the company’s commitment to CSR, by expressing the company’s commitment, by presenting others’ expectations about the company, or by providing supporting information and reasons in relation to the main claims made in the report (see Bondi & Yu, 2017).

Consider, for example, the words of an external expert.

(6) (Translation by Eni) “Talking about Sustainable Development today means embracing a new vision of the world around us [ . . . ] we must do so from these three angles [ . . . ] a big effort [ . . . ] is today more necessary than ever. And based on these three dimensions it is fundamentally important that the public and private spheres start to plan, co-finance and work together. The energy sector is ideal for this new approach [ . . . ] everyone should be aiming at and the steps required to realize that future. In the Post-2015 era public awareness will be essential to keep the attention focused on the Goals [ . . . ].” (Adapted from the lecture by Prof. Jeffrey D. Sachs at the Eni Enrico Mattei Foundation, titled “Feeding the Planet—Energy for Africa,” December 6, 2013; Eni 2013) (see Appendix 3 for original version in Italian)

Example 6 is the voice of a professor: In terms of pragmatic function, rather than simply expressing an independent evaluation, he gives suggestions on what should be done for appropriate CSR practice, dealing with the issue of “sustainable development.” His words appear directive in nature (in terms of speech act theory), as acts aimed at getting the addressee to adopt a specific behavior (Searle, 1969, p. 66). The directive nature of the quotes is evidenced by the extensive use of deontic expressions (we must, necessary, fundamentally important, everyone should, the steps required, will be essential to).

If we look at the quote in context, however, we can see that the necessity of certain actions may then serve as an argument supporting the legitimacy of the company, that is, it provides the company’s motivation for the execution of a certain action. Recontextualized in the CSR report, the voice is not only used to suggest that the company’s actions are right and legitimate, but more important, it is also used to imply that the company’s actions are useful to society and to meet societal needs, and therefore, the company is responsible and good. The quotation provides endorsement as well as legitimation. From an argumentative perspective, the voices of experts act as endoxa, “opinions that are accepted by everyone or by the majority, or by the wise men” (Rigotti & Greco Morasso, 2010, p. 501), working as major premises in the inferential structure of the argument. The important distinctive feature of these voices is that they do not address the company’s performance directly: They rather provide general authoritative guidance in the field, that is, the general rules according to which the company’s performance proves to be positive. The positive appraisal is left to the inferential (syllogistic) work of the reader on the basis of the general statement in the quote and other specific statements presented in the running text, typically showing that the company complies with the rule.

As shown above, voices from non-stakeholders either mainly appraise the company (and the company’s performance) in explicit endorsements or provide general rules acting as a premise leading inferentially to a positive appraisal of the company’s actions, on the basis of recommendations of an external party without any interest at stake.

Representation of Sources in CSR Reports

This section considers how the sources of voices are represented in the reporting discourse, following the framework proposed by Van Leeuwen (2008).

In our corpus, most direct quotations are presented with specified and determined sources. In terms of specification, a source is often represented as individuals (individualization) or in groups (assimilation). However, in our corpus when the sources are specified as groups, they are always represented in collectivation and are never treated as “statistics” and represented in aggregation. When the sources are determined, they are represented with names (nomination), or with roles, identities, and functions (categorization).

In most of the cases, the sources are both individualized and nominated, such as the following:

(7) “Ensuring entrepreneurs and small businesses have access to affordable financing is vitally important [ . . . ].” (Keri Gohman, Executive Vice President, Capital One Small Business Banking (Capital One 2013)

(8) “This is very important in the desert southwest,” says Paul Marken, completions supervisor. (ConocoPhillips 2013)

In Examples 7 and 8, the sources of voices are represented as individuals and nominated, that is, identified by full names and professional roles.

In some cases, although much less frequently, the sources are represented as groups (collectivation) and categorized according to the “identities and functions they share with others” (Van Leeuwen, 2008, p. 40), such as the following:

(9) (Authors’ translation) The employees who went to Tibet and Xinjiang to provide help stated, “Although the environmental conditions of the highland are arduous for those who always work here, these difficulties do not count, and we do not have any reason to be frightened either.” (ABC 2013) (see Appendix 3 for the original version in Chinese)

In the example above, the sources are represented as groups and categorized according to their roles (employees). Although readers are not able to identify who exactly said those words, the specification of the roles has some effects in emphasizing the company’s performance from a particular perspective.

Apart from the afore-mentioned representation types, scarce instances of genericization and indetermination are also found in our sample. Consider, for example, Figure 1.

Figure 1 illustrates the voice of stakeholders, which are represented in general “classes” (genericization) and in anonymous and unidentifiable images (indetermination). Strictly speaking, this kind of voice cannot be regarded as a “real” direct quotation, because most probably the words are written by the reporter instead of directly quoted from another source. However, it is a rhetorical choice of the reporter to present the words in the form of a direct quotation and to represent these as stakeholders’ concerns. Placed at the beginning of the report, these words have some rhetorical effects in demonstrating the company’s awareness of stakeholders’ needs and concerns.

The overall trend of the representation types in our corpus is illustrated in Table 3.

Table 3.

Representation Types in CSR-ICU.

	Frequency	Percentage
Genericization	11	2
Individualization	391	83
Collectivation	39	8
Aggregation	0	0
Other	31	7
Indetermination	10	2
Nomination	382	81
Categorization	49	10
Other	31	7

Note. CSR-ICU = corporate social responsibility, I for Italy, C for China, U for USA.

The results show that most of the sources are represented as single individuals (83%), while only a very limited number are represented as collective groups (8%). Still less used is genericization (2%). No instance of aggregation is found in our corpus (0%).

The dominance of individual voices can be attributed to one important feature of direct quotations: They are supposed to render the content of what the source has originally said (Coulmas, 1985), and the reporter “commits himself to repeating the actual words spoken” (Leech, 1974, p. 353). This commitment makes it easier to quote a single individual’s voice than a single voice from several individuals (i.e., collective).

Although viewing direct quotation as “verbatim production” has been challenged by many scholars (Clark & Gerrig, 1990; Mayes, 1990), this view still has some ground, especially when quoting in writing. At least, direct quotations preserve better the original formulation of the quoted voice in comparison with indirect quotations. When quoting voices from a group, normally it is easier to use indirect reports. Example 10 provides a summary representation of what the group said.

(10) Participants described the workshop as a practical learning experience for creating inclusive teams. (Citi 2013)

Even if collective quotes are rare, we still notice limited instances in the CSR reports, such as Example 9, presenting the collective voice of employees. In this case, although the quotation may come from only one employee’s original words, it is used to represent also the opinion of others.

In terms of indetermination and determination, the data show that only 2% of the instances are indetermined. Most of the sources are determined clearly by name (81%), while very few are determined only in terms of their role or occupation (10%). We also noticed that most of the sources represented in roles without determined names are employees or clients (see Example 9), while almost all managers, authorities, and experts are individualized and determined both by names and by professional roles. Since employees or clients are normally considered as social actors who do not have high authority and whose names are unknown to the targeted reader, it may be regarded as feasible to exclude this information (even if this reduces credibility). Although the categorization of source in roles somehow enhances the representativity of the voice shared by a group, we cannot ignore the fact that the indetermination of source may create more space for companies to present their own words in the form of direct quotation.

Direct Quotations Across Cultures

This section presents a comparative study of direct quotations used in the CSR reports published by companies based in Italy, China, and the United States. First, we provide an overview of the use of direct quotations in the three subcorpora. Afterwards, we look at the source types and how sources are represented across the three subcorpora.

The Use of Direct Quotations Across Cultures

As previously shown in Table 2, direct quotations are most frequently used in the Chinese subcorpus (224 instances), although the number of its total tokens is the lowest (471,050 tokens) among the three subcorpora. Except for two CSR reports (ICBC and GDPD), direct quotations are found in all the other 18 Chinese CSR reports. This implies that the direct quotation seems to be a commonly used rhetorical device in Chinese CSR reports produced by banks and energy companies.

In comparison with the Chinese subcorpus, the English one is slightly larger (499,889). However, fewer instances of direct quotations are found (182). Moreover, no instance is found in seven CSR reports, among which six are from energy companies. This suggests that, in the United States, in comparison with banks, energy companies seem to be less interested in this rhetorical strategy when compiling their CSR reports.

On the other hand, the Italian companies, either from the banking sector or the energy sector, do not show particular interest in the use of direct quotations. Although the Italian texts are much more extensive (1,107,122) in comparison with the Chinese subcorpus and the English one, only 66 instances of direct quotations are found. Moreover, among the 20 Italian CSR reports, only 7 reports use direct quotations, while in the other 13 reports, not even one instance is found.

Preference of Source Types Across Cultures

As shown above, different cultures show various degrees of interest in the use of direct quotations. Moreover, when citing voices, the choice of source types may present cross-cultural variations as well. This section focuses on how frequently internal sources and external sources are cited in CSR reports, typically addressing multistakeholders.

Figure 2 presents an overview of internal, external, and unspecified sources in the three subcorpora. The data show that, in general, the Italian and American companies tend to cite more internal voices than the Chinese companies, which show a clearer preference for external sources.

Figure 2.

Internal, external, and unspecified sources across the three subcorpora.

With greater detail, Figure 3 illustrates the distribution of the different social and professional roles of external sources (manager, employee, other) and internal sources (authority, client, community, evaluator, expert, partner, celebrity, other).

Figure 3.

Social and professional roles across the three subcorpora.

The data show noticeable cross-cultural preference regarding particular source types. In terms of internal voices, the Italian and English CSR reports tend to cite voices from managers (42% and 57%) rather than from employees (0% and 2%), while the Chinese CSR reports present more voices of employees (21%) than managers (9%).

In terms of external voices, the Chinese CSR reports show a preference for the use of voices of authorities (18%) and clients (16%). The American CSR reports, on the other hand, are characterized by more voices from experts (27%). In comparison, the Italian sample shows no noticeable preference for particular internal sources.

In general, the results suggest that Italian and English companies seem to regard professional roles as more important factors in enhancing the credibility of voices, considering that they place more emphasis on managers’ or experts’ voices. Chinese companies, on the other hand, prioritize voices from employees that highlight their values and experiences related to the company, voices from clients who appreciate the company’s services or products, and commendatory words from authorities.

Representation of Cited Sources Across Cultures

In the section entitled “Representation of Sources in CSR Reports,” we have investigated how the sources of direct quotations are represented in CSR reports. This section looks at the representation from a comparative perspective. Figure 4 illustrates the percentage of representation types across the three subcorpora.

Figure 4.

Percentage of representation types across the three subcorpora.

The results show that, in all the three subcorpora, the dominant types of representation are individualization and nomination. A noticeable culture-specific particularity is that sources cited in the Chinese sample are more frequently presented in groups (collectivation, 17%) and categorized by roles without specified names (categorization, 21%), in comparison with the Italian and American CSR reports. In Chinese, the sources presented in categories include mainly clients, employees, or local communities, while managers and authorities are almost always specified by names and roles.

Discussion and Conclusions

This section discusses the results in relation to the three research questions.

1. What types of textual voices are involved in CSR reports?

CSR communication is by nature polyphonic because it needs to address the relationship between the company and various stakeholders. Different voices can thus be found in the text, representing different people. As we observed in the analysis, one of the most important discourse functions of direct quotations is to endorse the company’s commitment to CSR and enhance their credibility, by presenting multiple voices beyond the predominant voice of the company.

Voices in CSR reports can be cited from internal sources, external sources, or unspecified sources. Internal sources mainly include managers and employees (who are also internal stakeholders). External sources include external stakeholders such as authorities, clients, local communities, partners, and so on, and non-stakeholders such as researchers/experts and evaluators. However, the apparent polyphony presented in the firm-centric CSR report tends to have a “symphonic” nature rather than a truly polyphonic one, because the company has the liberality to choose voices that are harmonious with the company’s stance. The most important voice is undoubtedly that of the company as a whole. Although the document is often a collective product of different people (e.g., consultancy company, CSR department of the company), in general it stands for the voice of the company, establishing an author-principal relationship (Goffman, 1981) between the company and the authorized writers, which actually realize the document.

This particular kind of “polyphony” does not have the effect of pushing forward the “communicative legitimacy” of the company (Castelló et al., 2013), which means embracing conflicts and dissent in corporate communication. On the other hand, the efforts of the company in presenting this pseudo-polyphony in CSR reports can be regarded as a direct or an indirect response to the multiplicity of issues raised: The presence of different textual voices does at least reflect the attention paid by the company to the role of some stakeholders. Moreover, if we keep in mind that CSR reports can be seen as “aspirational talk” (Christensen, Morsing, & Thyssen, 2013), that is, referring to future actions that may be the driving force for corporate effort and change, the positive comments in the direct quotations may have some prompting effect on the company’s commitment to maintain reality closer to what is said in the quotations.

2. How are different sources of voices represented in the CSR report?

In the CSR report, most of the quoted sources are represented as individuals with specified names, particularly those with higher administrative responsibility or expertise such as managers, authorities, and researchers. Much less frequent are the cases in which sources are categorized as “classes” or represented by groups with only specified personal role. This often happens when citing voices from employees, clients, or local community. Although most of the readers may have no idea about who the employees or clients are even when their names are specified, this difference in the choice of representation mode can somehow reflect the higher degree of “suppression” of certain groups of stakeholders that companies adopt when communicating CSR.

Moreover, the analysis has shown that different voices tend to take up different functions. If a feature shared by most voices is the endorsement of the company’s commitment to CSR, it is important to notice that this is realized by direct or indirect appraisal of the company for most stakeholders (with external stakeholders often also addressing their relation at stake with the company). Non-stakeholders, on the other hand, offer a more varied picture, with external authorities usually called in to provide data, general rules, and principles according to which the performance or commitment of the company can be inferentially judged as positive on the basis of the rules provided by the quote and the information provided in the running text.

3. What implications does the cross-cultural comparison provide on CSR discourse?

The comparative analysis confirms that different cultures are characterized by specific preferences regarding the use of direct quotations in presenting various voices in CSR reports. When considering the extent to which quotations are used, the Chinese CSR reports have a much clearer preference for direct quotations, and also the American CSR reports present a rather high frequency. The Italian CSR reports, on the other hand, show very limited interest in citing voices. These different tendencies in the use of direct quotations, which are rhetorical “designs” often presented in a visual way, are in line with the general style of the different CSR reports: The Italian samples seem to be more text-driven, while the American and Chinese CSR reports pay more attention to the multimodal aspects. Moreover, the noticeable use of direct quotations in Chinese CSR reports is evidence of a specific rhetorical convention in the Chinese culture: As exemplified in the studies of Matalene (1985) and Cai (1993), Chinese people seem to avoid expressing personal opinions and evaluation directly; instead, they prefer to refer to quotations, which may add weight (and credibility) to the opinions or ideas they are trying to express.

In terms of source types, Italian and American CSR reports show a preference for citing voices from managers. The Chinese CSR reports, on the other hand, present more voices from clients, employees, and authorities. Variations in the choice of source types may reflect culture-specific conceptions on the rhetorical effect of different source types in establishing credentials in reporting discourse. The dominant preference for managers’ voices in the Italian and American CSR reports, in comparison with the Chinese sample, may imply that, in the Italian and American business cultures, messages from persons with more administrative power or responsibilities within an organization are regarded as more rhetorically important in corporate communication with stakeholders. The Chinese business culture, on the other hand, tends to rely more on external voices, especially those from clients and authorities, to establish credentials and trust when communicating with stakeholders.

To sum up, what we are trying to suggest is that the CSR report, which aims at addressing various stakeholders, tends to present various voices in a symphonic way. This tendency is in line with the business nature of the genre and is identical in different cultures. However, cross-cultural variations are revealed by the specific choice of source types, as different business cultures may have different rhetorical conceptions of specific source types.

This study is limited in sample size and methodological approach. Only 60 CSR reports from two sectors were considered in our study. A more general picture on the business considerations of different voices in different national cultures can be provided if a wider range of CSR reports are included. Moreover, we investigated only direct quotations, and we did not consider the use of indirect quotations. Although we did not notice salient use of indirect quotations in the sample, the inclusion of this aspect can help improve the statistical integrity. For the methodological approach, we focused mainly on analyzing the texts, while interviews on how writers conceive of the presentation of various voices in CSR reports could make the results more comprehensive. Future study can expand the research scope and look at the relationship between source types and quotations’ discourse functions in relation to the surrounding text.

Supplemental Material

Online_appendicies – Supplemental material for Textual Voices in Corporate Reporting: A Cross-Cultural Analysis of Chinese, Italian, and American CSR Reports

Supplemental material, Online_appendicies for Textual Voices in Corporate Reporting: A Cross-Cultural Analysis of Chinese, Italian, and American CSR Reports by Marina Bondi and Danni Yu in International Journal of Business Communication

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

The online appendixes are available at .

Notes

ORCID iD

Marina Bondi

Author Biographies

Marina Bondi is a full professor at the University of Modena and Reggio Emilia, and director of the CLAVIER Centre (Corpus and LAnguage Variation In English Research). She has published extensively in the field of genre analysis, EAP, and corpus linguistics, with a focus on argumentative dialogue and language variation across genres, disciplines, and cultures.

Danni Yu is a lecturer at the Beijing Foreign Studies University. Her research focuses on corporate social responsibility reports, corporate communication and cross-cultural studies. She has published several papers in the field of CSR reporting and discourse analysis.

References

Bevitori

(2005). Attribution as evaluation: A corpus-based investigation of quotations in parliamentary discourse. ESP Across Cultures, 2, 7-20.

Blackwell

N. L.

Perlman

Tree

J. E. F.

(2015). Quotation as a multimodal construction. Journal of Pragmatics, 81, 1-7.

Bhatia

V. K.

(1993). Analysing genre: Language use in professional settings. London, England: Longman.

Bondi

Silver

(2004). Textual voices: A cross-disciplinary study of attribution in academic discourse. In Anderson

Bamford

(Eds.), Evaluation in spoken and written academic discourse (pp. 121-141). Roma, Italy: Officina.

Bondi

(2017). “I highly commend its efforts to ensure power supply”: Exploring the pragmatics of textual voices in Chinese and English. In Janebová

Lapshinova-Koltunski

Martinková

(Eds.), Contrasting English through corpora: Corpus-based contrastive analysis of English and other languages (pp. 218-247). Newcastle upon Tyne, England: Cambridge Scholar.

Bublitz

Bednarek

(2006). Reported speech: Pragmatic aspects. In Brown

Bauer

Berns

M. S.

Miller

J. E.

Hirst

(Eds.), Encyclopedia of language and linguistics (2nd ed., pp. 550-553). Amsterdam, Netherlands: Elsevier.

Cai

(1993, March). Beyond bad writing: Teaching English composition to Chinese ESL students. Paper presented at the College Composition and Communication Conference, San Diego, CA.

Calsamiglia

Ferrero

C. L.

(2003). Role and position of scientific voices: Reported speech in the media. Discourse Studies, 5, 147-173.

Castelló

Morsing

Schultz

(2013). Communicative dynamics and the polyphony of corporate social responsibility in the network society. Journal of Business Ethics, 118, 683-694.

10.

Christensen

L. T.

Morsing

Thyssen

(2011). The polyphony of corporate social responsibility: Deconstructing accountability and transparency in the context of identity and hypocrisy. In Cheney

May

Munshi

(Eds.), Handbook of communication ethics (pp. 457-474). New York, NY: Routledge.

11.

Christensen

L. T.

Morsing

Thyssen

(2013). CSR as aspirational talk. Organization, 20, 372-393.

12.

Clark

H. H.

Gerrig

R. J.

(1990). Quotations as demonstrations. Language, 66, 764-805.

13.

Coulmas

(1985). Direct and indirect speech: General problems and problems of Japanese. Journal of Pragmatics, 9(1), 41-63.

14.

European Commission. (2011). Communication from the commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions: A renewed EU strategy 2011-14 for corporate social responsibility (COM(2011) 681). Retrieved from https://publications.europa.eu/en/publication-detail/-/publication/ae5ada03-0dc3-48f8-9a32-0460e65ba7ed/language-en

15.

Freeman

R. E.

(1984). Strategic management: A stakeholder approach. Boston, MA: Pitman.

16.

Fløttum

Dahl

Kinn

(2006). Academic voices across languages and disciplines. Amsterdam, Netherlands: Benjamins.

17.

Fuoli

(2012). Assessing social responsibility: A quantitative analysis of Appraisal in BP’s and IKEA’s social reports. Discourse & Communication, 6(1), 55-81.

18.

Garzone

(2004). Annual company reports and CEOs’ letters: Discoursal features and cultural markedness. In Christopher

Gotti

(Eds.), Intercultural aspects of specialized communication (Vol. 14, pp. 311-342). Bern, Switzerland: Peter Lang.

19.

Glaser

Strauss

(1967). The discovery of grounded theory: Strategies for qualitative research. Chicago, IL: Aldine.

20.

Goffman

(1981). Forms of talk. Philadelphia: University of Pennsylvania Press.

21.

Gruber

(2015). Policy-oriented argumentation or ironic evaluation: A study of verbal quoting and positioning in Austrian politicians’ parliamentary debate contributions. Discourse Studies, 17, 682-702.

22.

Günthner

(1999). Polyphony and the “layering of voices” in reported dialogues: An analysis of the use of prosodic devices in everyday reported speech. Journal of Pragmatics, 31, 685-708.

23.

Hyland

(1999). Academic attribution: Citation and the construction of disciplinary knowledge. Applied Linguistics, 20, 341-367.

24.

Kuo

S. H.

(2001). Reported speech in Chinese political discourse. Discourse Studies, 3, 181-202.

25.

Leech

(1974). Semantics. Harmondsworth, England: Penguin.

26.

Leech

Short

(2007). Style in fiction. A linguistic introduction to English fictional prose. London, England: Longman. (Original work published 1981)

27.

Matalene

(1985). Contrastive rhetoric: An American writing teacher in China. College English, 47, 789-808.

28.

Mayes

(1990). Quotation in spoken English. Studies in Language, 14, 325-363.

29.

Petrić

(2007). Rhetorical functions of citations in high-and low-rated master’s theses. Journal of English for Academic Purposes, 6, 238-253.

30.

Rigotti

Greco Morasso

(2010). Comparing the argumentum model of topics to other contemporary approaches to argument schemes: The procedural and material components. Argumentation, 24, 489-512.

31.

Rosier

(2008). Le discours rapporté en Français. Paris, France: Ophrys.

32.

Searle

J. R.

(1969). Speech acts: An essay in the philosophy of language (Vol. 626). Cambridge, England: Cambridge University Press.

33.

Semino

Short

(2004). Corpus stylistics: Speech, writing and thought presentation in a corpus of English writing. London, England: Routledge.

34.

Shukrun-Nagar

(2009). Quotation markers as intertextual codes in electoral propaganda. Text & Talk, 29, 459-480.

35.

Sinclair

(1988). Mirror for a text. Journal of English and Foreign Languages, 1, 15-44.

36.

Sleurs

Jacobs

Van Waes

(2003). Constructing press releases, constructing quotations: A case study. Journal of Sociolinguistics, 7, 192-212.

37.

Smirnova

A. V.

(2009). Reported speech as an element of argumentative newspaper discourse. Discourse & Communication, 3(1), 79-103.

38.

Strauss

Corbin

(1994). Grounded theory methodology: An overview. In Denzin

N. K.

Lincoln

Y. S.

(Eds.), Handbook of qualitative research (pp. 273-285). Thousand Oaks, CA: Sage.

39.

Strauss

Corbin

(1998). Basics of qualitative research: Techniques and procedures for developing grounded theory (2nd ed.). Thousand Oaks, CA: Sage.

40.

Swales

(1986). Citation analysis and discourse Analysis. Applied Linguistics, 7, 39-56.

41.

Swales

(1990). Genre analysis: English in academic and research settings. Cambridge, England: Cambridge University Press.

42.

Tannen

(2007). Talking voices. Cambridge, England: Cambridge University Press. (Original work published 1989)

43.

Thompson

(1996). Voices in the text: Discourse perspectives on language reports. Applied Linguistics, 17, 501-530.

44.

Thompson

(1991). Evaluation in reporting verbs used in academic papers. Applied Linguistics, 12, 356-382.

45.

Thompson

Tribble

(2001). Looking at citations: Using corpora in English for academic purposes. Language Learning and Technology, 5(3), 91-105.

46.

Toolan

(2006). Speech and thought, Representation of. In Brown

Bauer

Berns

M. S.

Miller

J. E.

Hirst

(Eds.), Encyclopedia of language and linguistics (2nd ed., pp. 698-710). Amsterdam, Netherlands: Elsevier.

47.

Tree

J. E. F.

Tomlinson

J. M.

Jr. (2007). The rise of like in spontaneous quotations. Discourse Processes, 45(1), 85-102.

48.

Vandelanotte

(2009). Speech and thought representation in English. A cognitive-functional approach. Berlin, Germany: De Gruyter Mouton.

49.

van Eemeren

F. H.

Grootendorst

. (2004). A systematic theory of argumentation: The pragma-dialectical approach. Cambridge, England: Cambridge University Press.

50.

Van Leeuwen

. (2008). Discourse and practice: New tools for critical discourse analysis. New York, NY: Oxford University Press.

51.

Yung

V. K.

(1996). The presentation of voice in Chinese and English newspapers in Hong Kong. Perspectives: Working Papers in English and Communication, 8(1), 64-96.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.83 MB