Reporting verbs as evidentials in research articles in English and Croatian

Bašić, Ivana. (2017). Reporting verbs as evidentials in research articles in English and Croatian. PhD Thesis. Filozofski fakultet u Zagrebu, Department of Linguistics.
(Poslijediplomski doktorski studij lingvistike) [mentor Žic Fuchs, Milena].

Preview

PDF (Croatian) Download (2MB) | Preview

Abstract

This thesis was motivated by an attempt to gain new insight into the research question of how knowledge is constructed in academic discourse, namely, what rhetorical strategies writers of research articles use in order to make readers (members of the academic community) accept their claims as credible and reliable. The original contribution of this tesis is that it takes the linguistic notion of evidentiality as the framework for studying knowledge construction in research articles, whereby evidentiality is broadly defined as lingustic ways in which writers of research articles signal how they know what they claim, what evidence they base their claims on, and how they evaluate those claims. Furthermore, typical rhetorical ways in which writers of research articles frame their claims, and the “types of knowledge“ conventionally associated with these linguistic structures, are considered as evidential strategies whose ultimate aim is to gain readers' acceptance of those claims. Since academic discourse is largely made up of reports of the writer's own and other authors' research efforts, reasoning and findings, the analysis focuses on linguistic expressions containing reporting verbs as an obvious signal of reporting. The reporting verbs in the texts of research articles are considered as evidentials; namely, by analysing the interrelation of the lexical meaning of a particular reporting verb, the syntactic frame in which it appears (its arguments and complements), together with the particular reporting perspective, as well as the discourse functions of the sentence in which it features in the text, we try to infer what evidential meanings are expressed in the analysed sentence, in particular what the source of information is and what the attitude of the writer is towards the presented information. At the same time, we try to arrive at general conclusions about what “types of knowledge“, what kinds of evidential meanings and what aspects of evidentiality are conventionally expressed in some typical communicative situations in research articles, namely in reporting on other authors' or one's own research efforts, processes and findings. The analysis in this thesis encompasses the complete texts of 165 research articles in 9 research disciplines, in two languages – English and Croatian. The contrastive approach was chosen in order to prove the hypothesis that the fundamental pragmatic functions of research article texts - constructing credibility and gaining acceptance for one's claims – are the same in all disciplines, regardless of differences in individual disciplinary and community cultures, which are the result of the differences in the nature of the studied research problems and research methodologies, as well as various traditions and “epistemologies“ characteristic of particular research disciplines and communities. A substantial part of this thesis is dedicated to providing the theoretical background to the research problem that the analysis focuses on. We considered this justified and necessary both because we used it as a means of laying the foundation for our analysis and providing the reasoning for the methodology we used, and because we wanted to give our contribution to and provide an incentive for future research into the phenomenon of evidentiality, which has been largely neglected in Croatian linguistics. In examining the research problem and setting up the methodology for our analysis we relied on the theoretical framework and findings of cognitive linguistics. In this framework the primary function of human language is to communicate meaning, and meaning is seen as a dynamic process of conceptualizing experience, where decoding meanings, i.e. understanding, is based on the process of conventionalization, which presupposes implicit cooperation between the participants of the communicative process in a certain community. We can therefore study conventionalized linguistic units and linguistic behavior in a certain community as a means of gaining insight into how that community conceptualizes reality. In accordance with this reasoning, we analyse the linguistic behavior of research article writers, namely structures involving reporting verbs in the texts of research articles, as both a signal of the writer’s subjectivity, i.e. the ways in which the writer construes the conceptual content of the utterance, and a reflection of the epistemologies of particular disciplines and communities, i.e the conventional ways in which the conceptualizers (the writer and readers of the text) construct and categorize knowledge. The text of this thesis is divided into ten chapters. Chapter one discusses the term evidentiality and the existing definitions of this phenomenon in linguistics by providing a critical overview of contributions to investigating evidentiality, starting from the first mention of the phenomenon in the works of American linguists and anthropologists studying American Indian languages (Boas 1911, 1947; Sapir 1911, 1921; Lee 1938; Swadesh 1939) and continuing to the more recent relevant contributions (Givón 1982; Chafe&Nichols 1986; Žic Fuchs 1988; Ifantidou 1994, 2001; Mushin 2001; Friedman 2003; Aikhenvald 2004). We critcally assess the “narrow“ and “broad“ approaches to defining evidentiality, with the former (e.g. Anderson 1986; Willet 1988; Aikhenvald 2004) treating evidentiality as an exclusively grammatical phenomenon, and the latter placing it in the wider domain of pragmatics and epistemology (e.g. Givón 1982; Chafe 1986; Žic Fuchs 1988; Mushin 2001). More specifically, the proponents of the “narrow“ definition believe that the term evidentiality should be used only for grammatically encoded ways of pointing to the source of information, whose motivation is purely to be precise, without pointing to the speakers attitude towards the information. Conversely, the advocates of the “broad“ definition believe that evidentiality encompasses both the source of information and attitude toward it (traditionally referred to as epistemic modality or epistemic stance), as these phenomena are complementary and inseparable. In this view, evidential uses of various linguistic structures reflect ”natural epistemologies”, i.e. the ways humans categorize knowledge. In our thesis we subscribe to the broad definition of evidentiality as a notional linguistic category. We define evidentials functionally, as linguistic expressions pragmatically capable of expressing evidential meanings, and emphasize the importance of pragmatics and context in analysing evidential meanings. Chapter two provides an overview of various attempts to determine what constitutes as evidential meanings (Weber 1986; Willett 1988; Wierzbicka 1994; Givón 2001), and introduces the key terms which are used in linguistics to describe evidential meanings, such as source of information/knowledge; type of evidence/mode of knowing; direct (sensory and non-sensory) evidentials; indirect (reported, inferential) evidentials. Chapter three discusses the relation between evidentiality and complementary linguistic phenomena which reflect “natural epistemology“. We discuss the phenomena of linguistic subjectivity, epistemic modality and hedging, which are fundamental for understanding the reasoning behind the methodology we use in the analysis of a corpus of research articles in this thesis. We introduce the concept of conceptualizer, a language user who actively engages in the process of meaning construal (Langacker 1987) and offer a possible approach to evidentials seen as the expression of the epistemological stance of a conceptualizer in a particular situation (Mushin 2001). In this theoretical framework evidentials are seen as indicators of a conceptualizer's subjective construal of a situation as they express the conceptualizer's perspective concerning the knowledge of information they express in an utterance, based on the pragmatic context of a particular utterance. With regard to the phenomenon of epistemic modality, it is suggested that both evidentiality and modality are based on the implicit understanding of conceptualizers, i.e. participants in a communicative situation in a particular community, that some propositions require evidential marking, as they are conventionally challenged, whereas others are not open to dispute, either because they are taken for granted or because they are epistemically marked as less probable. We present two approaches to evidentiality, the first placing evidentiality under the umbrella of modality (Bybee (1985), Palmer (1986), Willett (1988)), and the second taking evidentiality as the umbrella term for all phenomena related to expressing „attitudes towards knowledge“, regardless of their grammatical expression (Chafe 1986). Finally, we discuss hedges as markers of the writer's evaluation/attitude/stance in research articles (Crompton 1997; Hyland 1996, 1998, 2005). Both hedges and evidentials are seen as modes of expressing the writer's epistemological stance, which presupposes active involvement of both the writers and readers (conceptualizers) in constructing the meaning of a particular utterance in a particular context. In our analysis we treat hedges as functional evidentials, since their communicative function is to gain acceptance for the writer’s claims and prevent readers from disputing their claims. Chapter four discusses the notion of evidential strategy and gives an overview of the most frequently described evidential strategies in English. According to Aikhenvald (2004), evidential strategies are the extensions of evidential categories and forms which encompass evidential-like meanings. We give an overview of contributions presenting particular linguistic structures as evidential strategies (Anderson (1986), analysing verbs of perception as potential evidentials; Aikhenvald (2004), discussing verbs of perception complemented by that or – ing clauses; Ifantidou (1994, 2001), focusing on parentheticals and adverbials as evidentials; and Aksu-Koç&Slobin (1986), Du Bois (1986) and Aikhenvald (2004), looking at reporting as an evidential strategy. By critically assessing the contributions of other authors, in particular the work of Čulić Viskota (2008), we emphasise the importance of analysing evidentials (specifically, the reporting verbs in our analysis) in the context of the discourse function in which they appear in the concrete example analysed, taking into account the wider textual and discursive context of the utterance. Chapter five discusses the relation between language and culture, namely the possibility of using evidential linguistic behaviour of particular communities as a way of gaining insight into their cultures, more specifically the ways in which they conceptualize and categorize types of knowledge. We give an overview of contributions which highlight the relationship between evidentiality and culture in languages and cultures where evidentiality is grammaticaly coded (Hardman 1986; Schlichter 1986; Weber 1986; Wierzbicka 1994). This chapter also discusses disciplinary cultures and communication in the world of science, focusing on the genre of the research article as the typical way in which researchers communicate within their discipline. We present Hyland’s (2011) notion of academic writing seen as knowledge construction in a particular discipline and a particular community, with various rhetorical conventions used as a means of gaining acceptance for one’s claims. We also introduce the term high and low risk discourse functions (Kuo 1999; Tang&John 1999), which will be relevant for our analysis of research articles in the corpus, where we analyse the use of reporting verbs as functional evidentials in the texts of research articles in English and Croatian, reflecting the particular disciplinary epistemologies and community cultures. Chapter six focuses on reporting as an evidential strategy in the scientific discourse of research articles. For the purposes of our analysis, the category of reporting verbs is defined functionally, as verbs featuring in reporting clauses by means of which writers of research articles report on their own or other authors' reasoning, research activities and processes, as well as the results and conclusions of these activities. In chapter seven we examine the research problem in our analysis of reporting verbs in research articles in English and Croatian. We present the hypotheses and the aims of the analysis, outline the corpus and the reasoning behind the setup of the corpus, and the methods used in analysing the corpus. Our analysis of the textual corpus focuses on reporting verbs, which are considered prompts (Langacker 1987), i.e. points of access to wider encyclopaedic knowledge that allow the conceptualizer, who metally processes (cognitively construes the conceptual content) to combine them further with other lexical units. The central hypothesis is that due to its communicative functions the research article can be considered a genre and discourse type particularly suited for studying evidentiality. Since we define evidentiality broadly, as a linguistic means of expressing the source of information/modes of knowing, but also the conceptualizer's attitude toward the information, we take utterances involving reporting verbs to be expressions of the writer's epistemological stance in the texts of research articles. Accordingly, the rhetorical strategies that a writer uses are seen as evidential strategies deliberately chosen in order to make the readers (the disciplinary community that the text addresses) accept their claims as credible. These strategies vary from one disciplinary community to another, but the ultimate communicative function is the same – gaining acceptance for the writer’s claims. Therefore, evidentiality is used in our analysis as an umbrella term which covers all the typical ways of pointing to sources of information and writers’ attitude to information in research articles. Evidential interpretation is considered to be the result of pragmatic inference, based on the interplay of the linguistic form and the context in which it is used. Reporting verbs in the texts of the analysed research articles in English and Croatian are analysed as evidentials, as points of access to insight into various ways of expressing different kinds of evidential meanings in the texts of research articles. The aim of the analysis was to investigate how writers of research articles linguistically signal the status of the knowledge they present in their utterances, especially how they signal the evidence on which they base their claims. We try to determine what types of knowledge, what aspects of evidentiality and what evidential meanings can be interpreted from linguistic forms that are used in some typical communicative situations and discourse functions in research articles. The corpus consists of 165 research article texts (95 in English and 70 in Croatian). To make the corpus representative in terms of including texts from both ”hard” and ”soft” disciplines, we included texts from 9 disciplines – computer engineering, mechanical engineering, physics, chemistry, biomedicine, psychology, sociology, linguistics, and literature. The genre of research article was chosen as the typical way for scientists to communicate their research findings to their own disciplinary community, whose members are invited to assess and verify their claims. High quality publications were chosen by consulting informants for each discipline, and all research articles were chosen randomly from one or more issues of the chosen journal. We used the following databases for access to full texts of research articles: Science Direct, arXiv, PLOS, ACM Digital Library, Taylor&Francis Online Journal Library, Sage Journals, Hrčak. The search was limited to issues between 2008 and 2014, as we wanted to gain insight into current rhetorical practises in scientific discourse. We searched the texts in the corpus for reporting verbs by identifying all reporting clauses and classifying them according to the reporting perspective used in the particular utterance (reporting from the personal perspective of the writer (e.g. I assume, I will show, we have found; smatramo, pronašli smo, pokazat ćemo); reporting from the perspective of the presented research (this article considers, our study demonstrates, results show, analysis reveals; istraživanja su utvrdila, rezultati pokazuju, analiza je potvrdila); reporting from an impersonal perspective (has been shown to, was found; pokazalo se, pronađeni su, utvrđeno je), reporting from the perspective of other authors (X argues, X shows, as X demonstrated; X navodi, X je pokazao, kako je utvrdio X); reporting by using evaluative devices (hedges or boosters) to express the writer's stance (it should be noted, (…) might be considered; valja istaknuti, može se smatrati, potrebno je promatrati). By looking at the discourse function of reports in a particular utterance, we try to arrive at conclusions about the relation between the lexical meanings of the particular reporting verbs and their evidential implications expressed by using these verbs in particular contexts. Also, the findings point to the differences in the frequency of use of particular verbs. Chapter eight provides selected examples of textual analysis of concrete sentences from the corpus according to reporting perspectives and presents the conclusions of this part of the analysis. The chosen method of analysis proved efficient in that it provided us with an overview of discourse preferences in individual texts (and parts of the texts), which made it possible to compare the discourse preferences between various disciplines and the two language communities. This method also enabled us to see clearly which reporting verbs appeared most frequently in all perspectives, i.e. which verbs constitute the basis of scientific reporting. Finally, the analysis clearly points to the relation between particular reporting perspectives and types of knowledge which are conventionally communicated by using these perspectives in the scientific discourse of research articles. The findings indicate that reporting from the personal perspective is related to expressing the personal epistemological stance. It is used to present knowledge as personal experience. Such knowledge is usually acquired directly, typically by observing a situation and drawing inferences on the basis of that observation, by hypothesising, or in some other way. Evaluative devices (hedges and boosters) can also be considered markers of personal epistemological stance. By using these devices writers of research articles present knowledge based on their inferences about a situation. Reporting from an impersonal perspective and reporting from the perspective of the research presented in an utterance is related to the factual epistemological stance. By using these two reporting perspectives writers of research articles emphasize the objectivity of their claims by stressing the importance of research processes and methodologies, rather than persons who conducted the research. Finally, reporting from the perspective of other authors is related to expressing knowledge acquired indirectly, by means of texts written by persons of authority. Using this reporting perspective is used as one of the basic ways of cumulatively constructing knowledge within a discipline. Furthermore, quoting authors of authority and engaging in a dialogue with their texts is one of the essential strategies used in research articles to gain acceptance for one's own claims and construct one's own knowledge as credible and reliable. Our analysis revealed that the most frequently used reporting verbs in all the texts in the studied corpus, in both English and Croatian, are verbs that may be classified as research verbs, namely verbs whose lexical meanings indicate that scientific knowledge is based on the process of conducting and observing the research process, and drawing inferences on the basis of direct observation. The most frequently used verbs among research verbs are verbs such as show, observe, indicate, see; pokazati/pokazivati, prikazati/prikazivati, vidjeti, with concepts of 'looking' and 'seeing', related to directly visually acquired knowledge, featuring prominently in their semantic structures. By presenting knowledge as directly visually acquired, these verbs have the potential of being used as an evidential strategy which emphasizes the reliability of the presented claims. However, their full epistemological potential is realised only through the interplay of a particular reporting verb and its arguments, the reporting perspective and the discourse function of the utterance in the specific context in which it appears in the text of the research article. Chapter nine presents the findings and conclusions of the analysis of the corpus from the point of view of individual disciplines and languages. The analysis of the corpus has shown that the conventional ways of rhetorically constructing information in research articles clearly reflect different epistemological beliefs in various disciplines, which are based on different natures of the research problems, which, in turn, require different research methodologies. In the corpus of texts from the field of literature and linguistics writers make extensive use of reporting from a personal perspective and the perspective of other authors, in combination with reporting verbs of thinking and speaking. In these disciplines knowledge is constructed both as acquired directly, through observing and inferring, as well as indirectly, through reading and assessing other authors' texts, since these two evidential strategies are seen as equally important and efficient. Conversely, disciplines whose research problems require an experimental approach prefer the rhetorical strategies of combining research verbs with reporting perspectives that emphasize the role of research processes and methodologies rather than the role of the researchers. In our corpus such evidential strategies predominate in the texts from the fields of medicine, chemistry, psychology and mechanical engineering. Interestingly, texts form the fields of physics and computer engineering in our corpus prevalently use the personal perspective in reporting, and they abound with reporting verbs of thinking, evaluative devices, and expressions which involve the reader. With regards to the differences between the English and Croatian corpora, Croatian texts largely prefer impersonal reporting, whereas personal perspective is prevalent in most discourse functions in English texts, with the exception of reporting on research methods, where impersonal reporting prevails in English as well. Furthermore, in personal reporting in Croatian authorial we is preferred in single-author articles, whereas in English first person singular is used in single-author personal reports. In both English and Croatian corpora reporting from the perspective of the research is the preferred strategy in the discourse function of reporting the results of one's own research. The differences in the evidential strategies used in the two corpora reflect varying cultural traditions in the two communities. Chapter ten provides general conclusions to the conducted analyses and critically assesses the original contribution of this thesis and its limitations, offering possible venues for further investigation of evidentiality in scientific discourse. The results of the analysis indicate that there are three types of knowledge conventionally constructed in the texts of research articles in the studied corpus: personal knowledge (expressed by reporting from a personal perspective and using typical evaluative devices, such as hedges and boosters), factual knowledge (expressed by means of reporting from an impersonal perspective and reporting from the perspective of the research), and knowledge based on the authority of reliable individuals (expressed by means of integral quotation, reporting from the perspective of other authors). All these ways of rhetorically constructing information in research articles are used as evidential strategies with the same goal – to enhance the credibility of the writer’s claims, prevent the readers from challenging them and ultimately to gain acceptance for them in the diciplinary community that the text addressess. The relevance and the scope of the findings of the textual analysis in this thesis are obviously limited to the scope of the methodology used in the analysis. Future research might focus on collecting quantitative data on the frequency of use of particular linguistic forms in a larger corpus of texts, which would lead to more conclusive findings concerning rhetorical preferences in particular disciplines. It is also necessary to focus on collecting data on the awareness and attitudes of members of various diciplinary communities to using various evidential strategies in their writing, as such findings are key to understanding the processes of becoming an efficient writer in a particular disciplinary community. Additionally, the findings of this research may be used as an incentive for a critical discourse analysis of the place and potential of scientific discourse as a discourse of truth and a discourse of power in today's post-truth society. Finally, the findings of the analysis in this thesis prove that reporting verbs are used as evidentials in the texts of research articles, and that their evidential interpretation is based on pragmatic inference, which confirms that evidentiality has to be investigated within the interpretative framework of pragmatics.

Actions (login required)

View Item