A Corpus – assisted Critical Discourse Analysis of Exclusion Strategies in Political Newspaper Articles: The March of Return in Gaza

This study examines if there is a favored linguistic exclusion strategy in Arabic and English newspaper articles reporting on the March of Return in Gaza and explores their potential impacts on the readers' minds. Ten newspaper articles were collected; five articles are in Arabic, each is from a different Arab country, while the others are the top-selling five newspapers in the UK. Data were analyzed by observing the linguistic structures and the representations of social actors from the CDA perspective adopted by Theo van Leeuwen and using two built corpora to calculate the frequencies of the exclusion linguistic structures found. The findings showed that Arabic and English newspaper articles used linguistic exclusion strategies differently to serve specific functions, such as: excluding the social actors involved in the case of the March of Return in Gaza, driving the reader's attention to the other social actor to view him as a victim or the party who abuses power, hiding or protecting the excluded social actor or driving the reader's attention to the event rather than the action.


Introduction
Political discourse is defined as the discourse practices in which politicians, organizations and citizens are engaged in the field of politics. When analyzing any political discourse, the focus is often directed to debates, speeches and hearings. As this paper takes exclusion strategies in political newspaper articles as the phenomenon of interest, it would be useful if a definition of exclusion is stated; Leeuwen( 2008 ) discussed the dichotomy of inclusion/exclusion and stated that exclusion is an important aspect of critical discourse analysis which is divided into two subcategories: total and partial exclusion. He used the terms radical and less radical exclusion as well. The first subcategory leaves no traces in the representation, excluding the social actors and their activities, while the latter falls into two subdivisions: suppression and backgrounding. Suppression means that there is no reference to the social actor in the text, while backgrounding means that the excluded social actor in a specific activity pops up later in another part of the text. This research answers the following questions : ( 1 ) which exclusion strategy is favored by the newspaper articles in both languages? ( 2 ) what are the functions of those strategies? ( 3 ) how are these strategies shown in language?
The study is qualitative and quantitative in nature; a linguistic observation of the structures of clauses and sentences is conducted in addition to an analysis of the representations of the social actors from the perspective of CDA set by van Leeuwen to explore the functions of exclusion strategies. The analysis also includes the use of two built corpora to calculate the frequencies of exclusion linguistic structures.
This paper tries to prove whether there is a favored exclusion strategy used by Arabic and English newspaper articles that reported on the case of the March of Return in Gaza and to illustrate the potential functions of those strategies on the readers' minds by describing how such strategies are manifested linguistically.
The data consist of ten newspaper articles reporting on the case of the March of Return in Gaza. Five Arabic newspapers articles are from Jordan, Lebanon, the UAE, Egypt and KSA. The other five articles are from the top five selling newspapers in the UK; The Independent, Metro, Daily Mail, The Daily Telegraph and The Sun. While many studies have been previously conducted to study exclusion strategies in many genres, this paper looks at the text from a different angle in that it conducts a contrastive analysis between texts of the same genre but with different languages. In addition, the study investigates exclusion strategies adopted by Arabic and English newspaper articles when reporting on the case of the March of Return in Gaza with no reference to inclusion strategies.

Literature review
Many studies have emphasized that the investigation of social actors' representations is a good starting point in searching for patterns of unquestioned meaning and ideological aspects of news discourse. This literature review sheds light on the studies that tackled the strategies of representing social actors and their manifestation in language.
Leeuwen (2008), in his book Discourse and Practice, presented a framework for analyzing how the participants of social practices can be, and are, represented in English discourse. In his chapter Representing Social Actors, he outlined the social and critical import of the categories of this framework and detailed the realization of each category. He also discussed the representations of social actors as groups ( assimilation ) and as individuals ( individualization ) and the ways in which actors can be categorized ( e.g. functionalization and relational identity ). In fact, analyzing the representations of social actors is seen as one of the core concepts of political discourse as in political debates or conflicts, there are normally two or more parties each undertakes a distinctive social or political action, and each action should be interpreted in order to show how power is used or abused.

Mehrpour and Dashti (2017), in their article Representation of Social Actors in J.Krishnaurti and Alan Watts' Philosophical Speeches:
A critical Discourse Analysis investigated the representation of social actors in the lectures of two philosophers, namely J.Krishnamurti and Alan Watts. They analyzed the transcribed lectures of the selected philosophers and examined the representations of social actors with emphasis on frequencies of inclusion and exclusion and the morpho-syntactic mechanisms employed in these speeches to exclude social actors. Social actors were represented differently in some discursive features. One philosopher addressed a wider range of audience beyond those people who were physically present in front of him without assuming the role of a teacher and speaking as " we " to a " one ", while the other was more involved with " I " and " you ". Undoubtedly, this study clearly showed how the speaker's ideologies or attitudes towards social actors affect his linguistic choice on both morphological and syntactic levels. Employing such a linguistic mechanism gives the speaker more flexibility to include a wide range of audiences, not only those physically attending or to include only his listeners whom he aims to address.
Another study that approached social actors' representation from a detailed textual perspective is by Majid (2010). In his article Actor Description, Action Attribution and Argumentation: Towards a systematization of CDA analytical categories in the representation of social groups, he attempted to make explicit various levels of discourse analysis on the representation of social groups, from detailed textual analysis to discourse topics to show how micro-level analytical categories are related to macrostructure within various levels of contexts. He suggested a three-level analytical framework for textual analysis of the representation of social groups, which divides text analysis into the domains of social actors, social action and argumentation. In this way, he looked at what is (not) in the text in terms of the three domains mentioned and investigated how these domains are linguistically realized through a set of linguistic processes. Sahragad and Davatgarzadeh ( 2010 ), in their study The Representation of Social Actors in Interchange Third Edition Series: A critical discourse analysis approach was used to investigate the linguistic representations of male and female social actors and the construction of gender identities in the Interchange, third edition. Forty-one reading passages were analyzed using van Leeuwen's (1996) framework and Halliday's transitivity model (Halliday and Matthiessen, 2004). All sentences were analyzed according to the two models previously mentioned, and the following elements of van Leeuwen's framework (1996) were considered to function as the criterion for the analysis: Inclusion, Exclusion, Activation, Subjection, Beneficilization, Participation, Possessivation, Circumstantialization, Functionalization, Classification, Relational Identification, Formalization, Semi-formalization, Informalization, Indetermination, Abstraction, Objectivation, Genericization, Individualization and Collectivization. The study was motivated by the fact that semantic and syntactic choices one makes in order to communicate serve to manifest their positioning and are based on the belief that one organizes their discourse in line with how they perceive a situation and the meanings they wish to convey (Halliday and Matthiessen 2014: 217). The study showed different representations of social actors in that females were portrayed as more prominent, successful, active and independent compared to males. The study challenges the traditional views which exclude the value of women in society. Koller (2009) put the emphasis on social actors and contexts with the aim of analyzing collective identity in discourse. She stated that the concept of social actors used is "indebted to van Leeuwen's ( 1996 ) taxonomy of social actors representation and the concept of context from Fairclough's ( 1995Fairclough's ( , 2003 notion of text as being embedded in the contexts of its production, distribution and reception as well as wider social, political, economic and cultural contexts. A micro-level analysis was conducted, and the focus was directed to the most prominent social actor incorporated into the text; the reader referred to the second person (you). The second person pronoun was used 27 times throughout the text. The direct address was found in statements, rhetorical questions and imperatives. This analysis calls for the universality of those findings into question, suggesting that through social actor representation, authors can construct outgroups in more detail than in-groups.

Critical discourse analysis
Scholars working under the umbrella of CDA see CDA from many perspectives. It is defined as ' a type of discourse analysis research that primarily studies the way social power abuse, dominance, and inequality are enacted, reproduced, and resisted by text and talk in social and political contexts (van Dijk, 2001, p.352). CDA looks into institutional, political, gender, and media discourses (Wodak, 2001) and how certain social groups may be ill-represented or misrepresented in various types of discourse.
Rendering to Dijk (2004), CDA is all about the analysis of underlying ideologies that, in turn, take part in the reproduction of or resistance against dominance or inequality. He pointed out that the focus of CDA is actually to criticize how power is abused. He also highlighted that dominance and inequality could elevate someone or a group that has the power as a dominant group and put the other group which has no or less power in an inferior position.

Theo Van Leeuwen's critical discourse model
Leeuwen's CDA model is primarily directed to show how people or groups are represented in the news in addition to representing one part in a higher position and the other in a lower position.
Leeuwen (2008) stated that representing social actors in the news is achieved through two analytical categories; inclusion and exclusion. Inclusion is putting the actors in the center of attention by being presented in the text clearly, while exclusion is omitting actors of social action by employing some linguistic mechanisms. He mentioned that exclusion had been seen as an important aspect of critical discourse analysis, particularly in political discourse (Leeuwen, 2008, p. 28). Thus, analyzing exclusion is more useful. As this paper is restricted to the study of exclusion, the linguistic mechanisms through which exclusion is manifested are to be mentioned.
Exclusion is divided into suppression and backgrounding. Suppression is the case in which social actors are totally absent from the text. This analytical category is realized linguistically through the following linguistic variations: Non-finite clauses; one famous example of non-finite clauses is infinitival clauses, passive agent deletion or passivation and nominalization; the use of a word that is not a noun as a noun .
Backgrounding is when social actors are not mentioned but can be inferred from the text or context. According to van Leeuwen ( 2008 ), excluded social actors pop up later in another part of the clause, sentence or text. Linguistic realizations of backgrounding can exist in a number of forms: simple ellipses in nonfinite clauses with -ing and -ed participles, infinitival clauses with -to and paratactic clauses; the reader knows the specific actor in the second clause of the sentence as in " Ahmad invited them to the ceremony, but the Wilsons didn't show up ". The following figure is the complete divisions of the Social Actor Network (SAR) proposed by Van Leeuwen (2008: 52) .

Data analysis
To find which exclusion strategy is favored in newspaper articles of each language, the newspaper articles were read thoroughly to comprehend the content. The focus was directed towards the clauses and sentences related to the representations of social actors, exclusion in particular. Then, the ten newspaper articles, including the headlines, were analyzed depending on the subdivision provided by van Leewen's(2008) model; exclusion is shown through suppression and backgrounding. Next, the actors involved in the text were classified. The case in hand is The March of Return in Gaza, and the two main actors are the Palestinian demonstrators and the Israeli forces; putting in mind that there might be other minor social actors who took part in the event evoked in the text; such as the Palestinian and Israeli presidents and the spokesman of the Palestinian Ministry of Health. These were included within the main social actors.
By observing the linguistic structures of the clauses and sentences related to the representation of the social actors in the ten newspaper articles, percentages for both suppression and backgrounding were calculated.   (2) shows that there were slight differences between the frequencies of exclusion strategies used in the Arabic newspapers. Suppression was more frequent in the text when excluding the Palestinians, while backgrounding was used more frequently when excluding the Israeli Army Forces.

Social Actors
To investigate the potential functions of exclusion strategies on the readers' minds, the representations of social actors were analyzed from the perspective of CDA put by van Leeuwen to explore the functions of exclusion strategies.

The Palestinians Suppression
To drive the reader's attention to the event rather than the action. Table 3: The potential functions of the exclusion strategies.
As shown in the previous table, exclusion strategies served to achieve four main functions. First, to drive the reader's attention to the other actor as a victim or an actor who abuses power. Second, to hide the excluded social actor. Third, to protect the social actor. Fourth, to draw the readers' focus to the event rather than the actors.
To examine how exclusion strategies are manifested in language, Sketch Engine was used. Sketch Engine is an ultimate corpus tool to create and search text corpora. Two corpora were created in both Arabic and English to calculate the frequencies of the linguistic exclusion structures. The first column to the left shows the two corpora of Arabic and English newspaper articles implemented in the study. Each in the second column was given a distinct name to process data easily. The third column shows the word count of each corpus.
Concerning the English corpus, fifty-two linguistic items were found and given tags and frequencies, as illustrated in figure  3 below . Ellipses tagged as ( PDT ) occurred only once, as in the following example.
Correspondingly, Linguistic features were also observed in the Arabic corpus. Twenty-nine linguistic items were found and given tags and frequencies as follows : To sum up, English newspaper articles were represented linguistically by five structures: nominalization, passivation, non-finite clauses, infinitive clauses and ellipses, with infinitive clauses achieving the highest frequency. Arabic newspaper articles were manifested linguistically by four structures: nominalization, passivation, non-finite clauses and ellipses, with nominalization achieving the highest frequency.

Results and discussion
This study was adequately powered and with a diverse sample of data to achieve the aim of determining favored exclusion strategies in Arabic and English newspaper articles reporting on the March of Return in Gaza, exploring exclusion linguistic structures implemented and their potential impacts on their readers.
The detailed analysis of the linguistic structures and the representations of the social actors involved in this political discourse revealed some interesting results. Linguistic exclusion structures such as nominalization, passivation, non-finite clauses, infinitival clauses and ellipses were found in different frequencies, as illustrated in the figures above; such apparent differences in the employment of exclusion strategies uncovered how injustice and power abuse are embedded within political discourse. For instance, in English articles, the reader could easily infer and find Palestinians as they were suppressed less than the Israeli Forces. However, in Arabic newspapers, they were suppressed more than the Israeli Forces, making it hard to identify.
In accordance with the linguistic variation found, a number of potential functions were achieved, such as driving the reader's attention to the other social actor, whether to view him as a victim or as the party which abuses power, hiding or protecting the excluded social actor or driving the reader's attention to the event rather than the action.
The findings can serve as a step toward understanding how journalists might place the same social actors in different positions.

Conclusion
Investigating the data from the perspectives of CDA and van Leeuwen's models proved to be helpful in understanding how someone's or group's position could be marginalized in text and in affirming that newspaper discourse is not neutral in presenting social actors in political discourse. For this, the present study mainly attempted to ensure if the social actors involved in the conflict reported were represented and positioned differently and to find out the potential impacts on the readers' views by investigating the favoured exclusion strategies along with their functions and linguistic manifestation in both languages. The study concluded that exclusion mechanisms were evidently manipulated in Arabic and English newspaper articles to exclude the social actors involved differently .This could act on the readers' attitudes in divergent ways in that the same social actor might be for once positioned as the party who abuses power or the one seen as the victim; in other cases, it might be completely hidden to have readers' attention directed to the event rather than the actors.
This paper contributes to the previous literature carried out in the field of critical discourse analysis by supporting the concept of positioning social actors in media through linguistic means to deliver the intention of the writer or speaker. More patently, this study contrasts Arabic and English newspaper articles reporting on the March of Return in Gaza and explores how linguistic exclusion strategy variation pictured the social actors. Thus, the findings can be regarded as a reference to journalists interested in the linguistic choice of newspaper articles to place the social actors in a way that corresponds to their attitudes or the overall political or historical convictions. For further research, it is recommended to search for more data; having a larger and more variant sample might shed light on the exact exclusion strategies favored in both Arabic and English newspaper articles resulting in getting more comprehensive generalizations on exclusion linguistic representations implemented.