British parliamentary speeches significantly influence public opinion and policy. Discourse analysis of these speeches provides valuable insights into economic policy formulation. However, traditional discourse analysis faces criticism for potential bias in data selection. This study bridges this gap by integrating corpus linguistics (CL) and discourse analysis (DA) methods. The research uses a Corpus-Assisted Discourse Study (CADS) approach to objectively analyze a large corpus of speeches (approximately two million words) delivered by Conservative and Labour MPs between 1900 and 2020. The study addresses the limitations of both CL and DA separately by employing a triangulatory method involving a back-and-forth process between quantitative CL and qualitative DA. The main research questions are: How can the synergy between CL and DA enhance data selection and representation? And how is the British economy discursively represented by British parliamentarians? The study aims to provide a more objective and representative analysis of the discourse surrounding the British economy.

The literature review highlights criticisms of both CL and DA when used independently. CL is criticized for decontextualizing linguistic features, introducing bias in data selection (e.g., through pre-selected seed words), and employing arbitrary cut-off points for keyword selection. DA, conversely, is criticized for its potential for fragmentary and unsystematic analysis and subjective interpretation due to limited data representativeness. The study argues that combining CL and DA can address these limitations, with CL providing systematic data selection and analysis, and DA offering contextual interpretation. Previous CADS studies demonstrating successful integration of these approaches are cited to support the study's methodology.

The methodology utilizes a five-step CADS procedure combining quantitative CL and qualitative DA. Step 1 (Seed Words Formulation): Three online dictionaries were consulted to create a comprehensive list of British Economy (BE) related terms (synonyms of 'poverty' and 'poor'). Duplicates and compounds were removed, leaving 435 single-word synonyms. Step 2 (Examining the Corpus for British Economy Seed Words): The 435 seed words were searched in the corpus. Words occurring ≥10 times (a standard cut-off point) were selected for further analysis (93 words). This step employed AntConc software for keyword identification and concordance extraction. Step 3 (British Economy Keywords): Keyword analysis was conducted comparing the corpus to two reference corpora: the British National Corpus (BNC) and the CORpus of Political Speeches (CORPS). This identified keywords significantly different from general English and political discourse corpora, further reducing the number of terms for qualitative analysis. Step 4 (Meaningful Keywords of the British Economy: KWIC Analysis): A KeyWords In Context (KWIC) analysis was performed examining the context of each keyword to filter out those thematically irrelevant to the BE (false positives). This qualitative examination focused on 5-word span collocates and Mutual Information (MI) scores to ensure contextual relevance, resulting in a final list of 28 KWICs. Step 5 (Discourses of British Economy): Word sketch analysis (using Sketch Engine) examined the collocational behavior of the 28 keywords. Discourse types were then identified based on the semantic categories of the keywords' collocates and grammatical relations. This process identified four broad discourses: Finance, Workforce, Living Standards, and Hardship.

The analysis of the two million-word corpus of British parliamentary speeches from 1900–2020 revealed that the dominant discourses concerning the British economy were finance and hardship. The study utilized a five-step methodology integrating corpus linguistic (CL) techniques and discourse analysis (DA) to produce a representative corpus for qualitative investigation. Three key semantic categories, namely alleviation, scale, and source, emerged from analyzing the collocational behavior of keywords. A total of 28 keywords related to the British economy were identified and categorized into four discourses: finance, workforce, living standards, and hardship. Finance and hardship discourses were significantly more prevalent than those relating to workforce and living standards. The analysis of the linguistic behavior of keywords further revealed patterns of action to be taken, scale, and the source of economic issues as recurring themes in both the Conservative and Labour corpora. Differences in keyword frequency and usage patterns were observed between the two parties, suggesting variations in their framing of economic issues.

The study's findings demonstrate the effectiveness of integrating CL and DA for a more objective and representative analysis of political discourse. The combined approach successfully addressed the limitations of each method when applied separately, resulting in a richer understanding of the discursive representations of the British economy. The prevalence of finance and hardship discourses highlights the central role of financial stability and social welfare in shaping political discourse. The identification of semantic categories (alleviation, scale, source) sheds light on how economic issues are framed and addressed in parliamentary debates. This integrated approach provides a valuable framework for analyzing political discourse on complex socio-economic issues.

This study successfully demonstrated the synergy between corpus linguistics and discourse analysis, enhancing data selection and representation for a more objective analysis of the British economy in parliamentary speeches. The five-step methodological procedure yielded four main discourses: finance, workforce, living standards, and hardship, with finance and hardship being the most frequent. This integrated approach offers a valuable methodological framework for future studies investigating political discourse on socio-economic topics. Future research could explore public reception of these discourses to understand their impact on public opinion and policy.

While the study utilized a large corpus, it was limited to speeches from Conservative and Labour MPs, potentially overlooking perspectives from other parties. The temporal scope, while extensive, might not capture nuances of shorter-term economic shifts. The reliance on publicly available data sources could introduce biases present in those sources. Future studies could expand the corpus to include a wider range of political actors and explore different time scales to further refine the analysis.