Corpus analysis

Khosla 1 Pooja Devi Khosla Comparing the representation of the asylum seekers in The UK and The USA newspaper The Guardian. Introduction This study analyses the discourse of asylum seekers in the UK and the USA. In particular, this paper examines the representation of asylum seekers in the UK and USA newspapers: ‘The Guardian’. RASIM (Refugees, asylum seekers, and immigrants) has grabbed the attention of many researchers for Critical Discourse Analysis (CDA). Thus, many pieces of research have been conducted in this context; namely, Oregan & Riordan compared the representation of Refugees, Asylum Seekers, and Migrants in the Irish and UK Press, Majid KhosraviNik analyzed the Representation of Refugees, Asylum Seekers, and Immigrants in British Newspapers, Gabrielatos & Baker did a Corpus Analysis of Discursive Constructions of Refugees and Asylum Seekers in the UK Press and Majid KhosraviNik analyzed the Representation of Refugees, Asylum Seekers and Immigrants in British Newspapers, during the Balkan Conflict (1999) and the British General Election (2005). These researches indicate the negative stance adopted by media and the politicians towards the ethnic outgroups arriving at their nations. However, all these articles are related to before and during refugee crisis events so some changes might be expected in the representation of asylum seekers in recent years. The research question of this paper is to analyses the discursive structure of asylum seekers in the press and identify any differences in the representation of asylum seekers in the UK and

Khosla 2 USA media. This study aims to analyze major topics/issues related to asylum seekers and figure semantic prosodies narrated in the news articles regarding asylum seekers. Data and Data Collection Method The data for this paper Orange App was used which is an open-source data visualization and analysis tool in which data were collected through the search query: asylum seekers. Pre- processing of the news articles was done to remove the unwanted records from the data sets thus the only UK and USA news was selected. Then the selected articles were saved as Text files in the personal computer using Python Script. Data collection was done only from one type of source i.e. liberal newspaper which was the Guardian. To limit the data, only one newspaper was selected. Further, to validate the genuineness of the source of data collection three factors were examined thoroughly: first, online availability of the news articles, second: news articles are free of cost, and third: a significant amount of readers must be targeted by the newspaper. While distinct sources provided different numbers, but on average theguardian.com is visited by nearly 85 million visitors. Data consists of all the newspaper articles related to asylum seekers from the year 2018-2019. As the aim of this article is to analyze the discursive representation of asylum seekers in media and comparing their representation, dates were restricted from the year 2018-2019 because logically previous studies are focusing on the refugee crisis-era while variability will be analyzed in this paper. Consequently, USA and UK corpora comprised of 9,63,495 words was obtained. Limitations

Khosla 3 However, due to the limited number of articles used in the paper, it is hard to identify any shift in the pattern of the opinion of the editors over time. Also, small corpora may lack some of the features in focus (Baker et al. 285). Articles from only one source are gathered thus limiting the variability of the lexical patterns in the context. Further, the corpus emphasis on the opinion of the powerful spokesperson on behalf of asylum seekers rather than asylum seekers themselves. The corpus in this paper is restricted to the word asylum seeker(s) though some articles might not prefer to use word asylum seeker notably but might refer them with other rude words, for instance, aliens, illegals, or detainees. A small-scale analysis may not be able to identify which linguistic patterns are cumulatively frequent (and therefore likely to represent powerful discourses) and those which are less frequent (and therefore may constitute minority or resistant discourses) (Baker et al. 285). Methodology Keeping in mind the objective of the research to identify patterns in UK and USA coverage by the Guardian about asylum seekers, Corpus Linguistics (CL) techniques to do CDA was considered to be an appropriate methodological framework (Paul Baker et al. 283). The study starts with the corpus linguistics (CL), to get a wider categorization of the data through quantitative/statistical analysis which enables the further analysis of significant patterns in the corpus, while CDA like topics and semantic prosody lead to qualitative analysis which helps in mapping this pattern in context and leads to a deeper analysis of specific data (Oregan & Riordan, 748). In light of this, the analysis of the corpus is done using AntConc. In the corpus linguistics aspect of the study, the notions keyness and collocation were implemented. Also, other notions of CDA were employed, like concordance, topics, and semantic prosody. Frequency Analysis:

Khosla 4 In this paper, keyword analysis is used to identify the words which are significantly frequent in the UK and USA corpora. The categorization of keywords related to particular topics, viewpoints, and contrary topoi (Gabrielatos & Baker, 10). and investigating their usage in context made it possible to identify discourses of asylum seeker(s) in the UK and USA corpora. To identify the keywords to be examined, BNC wordlist was used as a reference wordlist, the threshold was set at the minimum value of p i.e. p<0.0001 which helped to restrict the number of keywords and enhanced the accuracy of the keyness of a word (Gabrielatos & Baker, 10)and keyword list was sorted by its keyness. Collocation Analysis Collocation is a method of interpreting associations and meanings between the lexical words because words hold meaning only by the context that they occur in. Therefore, collocation is done to get the information about the most salient points associated with a node (Stubbs, 1996) and to extract a semantic analysis of a word (Sinclair). In this paper, for collocation analysis, window span was set at five words left, and five words right of the node and collocability was decided using the measure: log-likelihood (LL) (Gabrielatos & Baker, 11). According to Gabrielatos & Baker, LL was used to remove the words which have a high probability of co-occurring with the nodes due to chance (11). Hence the keyword list was sorted by stat in order to discover the statistical significance of the co-occurrence. Concordance The collocational analysis can be further expanded using concordance analysis (Gabrielatos & Baker, 15) which is used for the ‘qualitative analysis’ and close examination of the corpus. Concordance provides the list of the appearances of a particular search term in the context which sometimes is referred to as KWIC i.e. keyword in context with a few words at the left and few words to the right of the KWIC.

Khosla 5 Findings and Discussion To identify the frequency of the lexical words thereafter to identify primary categorization for a thorough diachronic analysis of the most frequently appearing items, the frequency list is generated (Oregan & Riordan, 749). The first 20 keywords of the two corpora (Figure1) indicate that the USA newspaper revolves around political stance like democrats, republican, and Donald whereas UK newspaper primarily focuses on the asylum seekers like asylum, seekers, and immigration. Further, the use of words like federal and spokesperson in both the corpora indicates that some groups/ parties speaks on the behalf of these outgroups (asylum seekers) rather than the outgroup themselves. Figure 1 The further analysis of the next 20 keywords (Figure 2) was executed to identify more variations in the representation of the asylum seekers in the corpora. In the USA corpus, it was evident that some regions were specifically targeted like Texas and Washington while discussing in regards to asylum seekers and the continuous usage of term republicans and

Khosla 6 congress validated the political stance in this context whereas in the UK corpus major concern was the dismissal of the refugees indicated by the words like hostile, persecution and removal. Also, the UK corpus consists of words like hostile, persecution, removal, and deportation which portrays asylum seekers as victims. On the other hand, in the USA corpus, keyword ‘white’ points towards the discrimination and racism asylum seekers face in the USA. Figure 2 The collocates of asylum seekers gave an insight into the point of view of the newspapers while representing asylum seekers. For systematic analysis of discourse, Gabrielatos and Baker’s strategy of categorization was done for which 5 CDA-informed categories were formed represented in figure 3 (21). The most common collocate in both the corpora is refugees which are frequently represented along with the asylum seekers using and conjunction. However, the entirely

Khosla 7 different collocates in UK and USA corpora are ‘accommodation’ in the former and ‘returning’ in later. Figure 3 The most common aspect associated with the matter of asylum seekers was there number like ‘hundreds’ and ‘more’ in UK corpus (Figure 4) and ‘many’ and ‘thousand’ in USA corpus (Figure 5). Hence, some kind of quantification was observed in both the corpora while doing the concordance analysis of term asylum seeker(s) which sometimes represents asylum seekers as an alarming issue.

Khosla 8 Figure 4 Figure 5 Entry/exit and accommodation are highly frequent categories. Considering previous researches on asylum seekers it was expected that the asylum seekers are unwanted but unavoidable in countries like the UK and the USA. Collocates of ‘accommodation’ and ‘housing’ in UK corpus and collocates of ‘deportation’ and ‘returning’ in USA corpus, portray a completely different scenario which is an interesting variation in the collocates of both corpora. Plight category represents the problems that asylum seekers face when they come to this country. Though this category was not widely populated but highly collocated with the groups like young and children and only 28 concordances out of 422 were observed in UK corpus but surely built a sense of sympathy and compassion in the minds of the readers. However, in the USA corpus, few collocates associated with the sufferings of the asylum

Khosla 9 seekers were found and that too primarily targeted the politicians using asylum seekers as an issue shown in the example below. Help category describes the formal aid that asylum seekers get from the government representing them as tragic victims which are supported by the reoccurring pattern of ‘for asylum seeker(s)’ traces the ‘helping’ discourse which is quite frequent with 25 concordances in total 215 files analysis in UK context. Further construction of the ‘helping’ discourse is contributed by terms like help, concern, rescue, and shelter. However, a completely different scenario was observed in the USA context where ‘for asylum seeker’ pattern represented the miseries and obstacles asylum seekers face due to government policies. Subsidiary linguistics strategies of individualization of asylum seekers are implemented in the UK discourse by singling out the person or group by using their proper names like using noun ‘Macharia’ a famous Rugby player which was absent in the USA discourse. For politicians in the USA, the asylum seekers are a part of their semantic components. Republican, democrats all such words make asylum seekers as a matter which can be called as politician rivalry discourses in which negativization of asylum seekers is a part of criticizing the policies of the rival’s policies which was not the case Conclusion The study examined the representation of asylum seekers in the UK and USA media. Using Baker et al., a combination of CL and CDA was used to analyze the corpora from the collected articles from the USA and UK media, the Guardian. The findings indicate that

Khosla 10 overall, asylum seekers are just reduced to an issue and linguistically backgrounded and hence categorized as an outgroup over whom there is an infinite ongoing debate. Though there were articles that followed a sympathetic structure while representing asylum seekers, but the majority of the articles represented them as an unwanted group whose removal from their culture is a major issue. Tracing on the interdiscursive and intertextual elements of the former and adjoining reiterating topics, such as Windrush scandals in UK corpus and dehumanizing term in USA corpus, the examined result reflects a sympathetic structure, but this does not indicate that all articles have followed the same linguistics and discursive strategy. Overall, the asylum seekers are presented in less negative way in UK corpus as compares to USA corpus.

Khosla 11 Works Cited BAKER, PAUL, et al. \"A Useful Methodological Synergy? Combining Critical Discourse Analysis and Corpus Linguistics to Examine Discourses of Refugees and Asylum Seekers in the UK Press.\" Discourse & Society, vol. 19, no. 3, 2008, pp. 273-306. Gabrielatos, Costas, and Paul Baker. \"Fleeing, Sneaking, Flooding: A Corpus Analysis of Discursive Constructions of Refugees and Asylum Seekers in the UK Press, 1996- 2005.\" Journal of English Linguistics, vol. 36, no. 1, 2008, pp. 5-38. KhosraviNik, Majid. \"The Representation of Refugees, Asylum Seekers and Immigrants in British Newspapers: A Critical Discourse Analysis.\" Journal of Language and Politics, vol. 9, no. 1, 2010, pp. 1-28. KHOSRAVINIK, MAJID. \"The Representation of Refugees, Asylum Seekers and Immigrants in British Newspapers, during the Balkan Conflict (1999) and the British General Election (2005).\" Discourse & Society, vol. 20, no. 4, 2009, pp. 477-498. O'Regan, Veronica, and Elaine Riordan. \"Comparing the Representation of Refugees, Asylum Seekers and Migrants in the Irish and UK Press: A Corpus-Based Critical Discourse Analysis.\" Journal of Language and Politics, vol. 17, no. 6, 2018, pp. 744-768. Sinclair, John. Corpus, Concordance, Collocation. Oxford University Press, Oxford,1991. Stubbs, M. “Grammar, Text, and Ideology: Computer-Assisted Methods in the Linguistics of Representation.” Applied Linguistics, vol. 15, no. 2, Jan. 1994, pp. 201–223.