36th European Accounting Association

TRUST THE TEXT. EXTENDING METHODOLOGIES FOR THE ANALYSIS OF ACCOUNTING NARRATIVES WITH A CORPUS LINGUISTICS/CORPUS-DRIVEN APPROACH This paper, drawing on Sydserff and Weetman’s (2002) seminal work on the syntactic dimension of accounting narratives, addresses the call for more objective computerised methods, applications, and interdisciplinary approaches in investigating corporate narrative documents from a systematic, detailed linguistic perspective. Previous research appears to be one-dimensional in applying only a quantitative or qualitative analysis and ignoring the relationship between type of disclosure, topic and context. This paper, by contrast, extends the range of methodologies by introducing/illustrating the application of a Corpus Linguistics/corpus-driven approach integrated with discourse analysis. The automated nature of Corpus linguistics, its theoretical basis in contextual and functional theory of meaning, its balanced quantitative and qualitative analysis point to its strength in validity. Thus, in attempting to develop more objective, manageable and replicable computerised methods, this paper contributes to move beyond readability formulas, keyword dictionaries, indexicals through the study of recurring phraseology based on keyword frequency, concordances, extended units of meaning, semantic prosody and the bi-modal influence between context and accounting narratives. This innovative methodology is useful at both a managerial operational level and a tertiary educational level, while representing a step ahead called for by several researchers.