A corpus-based approach to corporate communication research

Abstract

Corporate Social Responsibility (CSR) has become a decisive reputation tool for companies and as such, a key concept in corporate communication as a phenomenon of intercultural and global significance. This has warranted a comprehensive examination of the language of CSR reports reflecting the principles of the corporate culture. Studies exploring the narratives of CSR reporting currently lack insights into the distribution of meaningful priorities evidenced in language use. This study sets out to explore the linguistic environment of the most frequently occurring language tokens to identify recurrent language patterns used to ensure efficient CSR reporting, and to further establish priority directions in CSR narrative composition evidenced in language use. A corpus-based approach and contextual analysis were adopted to examine CSR reports issued by Microsoft over the last seven years and recognised as an example of best practices in the corporate field. The corpus was compiled using the Prime Machine corpus concordancer tool and comprised 99,176 tokens. Following the study results, the study makes a number of inferences regarding the use of pronouns, “Microsoft + a verb denoting positive action”, “more + than,” “more + adjective”, “Corporate” as part of compound terminological units, as well as a set of key tokens encountered within a descriptive linguistic environment with positive connotation. This, in turn, proved helpful in identifying the hierarchy of priorities distribution revealed in the course of material analysis. The results contribute to a systemic appreciation of corporate language policies facilitating efficient stakeholder communication and can be used in further research investigating related matters of scientific interest.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 23 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.