Detecting Bias through Discourse Analysis

Adrienn Keszei

Detecting Bias through Discourse Analysis

Language constructs one’s world, and it has become one of the most important foundations of humanity, so much so that it is impossible to imagine life without the constant flow of communication. However, the language codes that particular groups of people (be it a nation or a smaller community) share are simply not sufficient in themselves to result in a successful exchanging of information. The vast amount of knowledge that human beings share with each other would seem incoherent and chaotic without context. 

Journalism, one of the most common means of sharing information on a national or global level, heavily relies on context, specifically on social context. In addition, the ‘messages’, which newspapers produce are constructed by the intertwined relationship between language, ideology and power, especially regarding the coverage of politics. According to Joseph (2006), “language evolved as an ultra-efficient means of distinguishing allies from enemies and of grooming allies and potential allies” (p. 1). Politicians have taken advantage of this persuasive power of language for centuries, but they are not the only ones who regularly employ rhetoric.

Primarily, “journalism exists to enable citizens to better understand their lives and their positions in the world” by reporting the undistorted truth (Richardson, 2007, p. 7). Nevertheless, newspapers, either on purpose or subconsciously, often use language that show bias by favouring one ideology or position over the other. In theory, journalists are supposed to be objective, but in practice one can easily identify signs that indicate subjectivity, or at the very least, the lack of balance. Richardson (2007) argues that while most people can recognize bias in newspaper articles, it is difficult to identify how the author expresses his or her own bias; however, the “interpretative, contextual and constructivist approach” of discourse analysis provides a helpful tool to better evaluate what lies beneath the news (p. 15). The purpose of this paper is to analyse two articles reporting on the same event, published by ideologically opposed Hungarian newspapers. Although the authors of these news articles are careful to avoid open bias, with the help of a corpus-based discourse analysis, it is easy to identify the key indicators of both newspapers’ political stance.

Bednarek (2006) introduced a new approach in order “to analyse the phenomenon of speaker opinion – variously known as evaluation, appraisal and stance within linguistics; these “evaluative parameters” serve to determine what and how is implied with the specific choices of vocabulary (p. 3). Evaluation is important because essentially every newspaper evaluates the published facts; otherwise news would be a list of data, rather than a narrative. Thus, by commenting on the information presented the authors reveal their own attitudes towards a situation. Appraisal and stance are the most significant indicators of the speaker’s evaluation, showing the bias in the written discourse. Stance if often “defined as the overt expression of the speaker’s attitudes, feelings, judgements, or commitment concerning his/her message, including the indication of the speaker’s degree of commitment towards the truthfulness of the message” (Bednarek, 2006, p. 25).

            After the 2014 Parliamentary Elections were held in Hungary, two prominent newspapers, Magyar Nemzet and Népszabadság, both published articles about the results. Magyar Nemzet is known to sympathize with Fidesz, the ruling party, and the corpus based examination of its article also supports this claim. The title immediately indicates praise “There has never been such a double victory” [Ilyen duplázás még nem volt], which is further confirmed by the constant and repetitive uses of phrases such as “landslide victory” or “overwhelming majority” regarding the consecutive victory of Fidesz (Z., 2014). The article also comments on the opposing party, revealing an unfavourable tone towards it: “the Left, comprised of discredited characters” or “the weakness of the opposition” (Z., 2014). According to Bednarek (2006), such positive and negative phrases are overt markers of stance (p. 26). While ‘overwhelming majority’ and ‘landslide victory’ signal a positive attitude, ‘weakness’ and ‘discredited characters’ are strong markers of judgement and prejudice. It cannot be coincidental that the negative markers are used in connection with the opposition, whereas the more positive lexical markers are used regarding the ruling party, implying the author’s ideological bias against the opposition.

            As Népszabadság is a publication closely related to MSZP (the opposition), bias against Fidesz is expected to appear in articles chronicling its victory. The title clearly implies the improbability of the Fidesz victory: “In a European democracy it would be impossible to imagine the Orbán’s two-thirds majority” ("Európai demokráciákban elképzelhetetlen lenne az orbáni kétharmad" 2014). The article called the overwhelming majority a case of “serious inequality”,  and further criticized the prime minister claiming that he “as usual forgot to give credit to the EU’s financial support and boasted about each renovated building as his own success” ("Európai demokráciákban elképzelhetetlen," 2014). Not only did the author imply bias with the tone of the article, but the lexical choices in the above sentence are overt negative markers. The phrase ‘as usual’ references earlier examples of the prime minister’s forgetfulness, relying on the readers’ contextual knowledge of previous events. Furthermore, the verb ‘boast’ has a negative connotation, thus it is most probably used in this context to elicit a negative emotional response. The above examples correspond with Bednarek’s (2006) “judgement system” that claims that media discourse often reveals speakers passing judgement, “morally evaluating human behaviour, by reference to a set of norms” (p. 28).

            The major objective of this paper was to assess newspaper articles based on their lexical and contextual background.  While this essay cannot provide a thorough analysis, the examples above all support the claims regarding each newspaper’s political stance. Therefore, the lexical choices in newspaper articles always need to be taken into account if one wishes to accurately identify bias. However, journalists also have to rely on their intended audience, more importantly, their social knowledge, that helps them understand the intended message. According to Richardson (2007), “we make sense of discourse partly by making guesses - usually unconsciously - based on social knowledge” (p. 23). Communication may be a delayed process when it comes to printed media, the readers can be regarded as active contributors all the same. Regardless of the masterful use of rhetoric and lexical content, to elicit the intended response, journalists also need the public to process the provided information applying their social, political, or economic knowledge.

Reference List

 

Bednarek, M. (2006). Evaluation in media discourse: Analysis of a newspaper corpus. New York: Continuum.

Európai demokráciákban elképzelhetetlen lenne az orbáni kétharmad. (2014, April 7). Népszabadság Online. Retrieved from http://nol.hu/kulfold/szlovakia-europai-demokraciakban-elkepzelhetetlen-lenne-az-orbani-ketharmad-1455209

Joseph, J. E. (2006). Language and politics. Edinburgh: Edinburgh University Press.

Richardson, J. E. (2007). Analysing newspapers: An approach from critical discourse analysis. (1st ed.). Houndmills: Palgrave Macmillan.

Z., V. (2014, April 8). Ilyen duplázás még nem volt. Magyar Nemzet Online. Retrieved from http://mno.hu/parlamenti_valasztas_2014/ilyen-duplazas-meg-nem-volt-1220290