A Two-layer Annotation Scheme for the Pragmatic Analysis of the Passive

Abstract Current pragmatic research shows that structures such as the periphrastic passive influence the interpretation of utterance meaning, yet none specifies how linguistic structure gives rise to a given pragmatic effect. The existing body of literature remains mainly theoretical and descriptive rather than methodological. This study thus offers a replicable double-layered annotation scheme linking observable linguistic cues such as verb type, lexical polarity, oblique complement, and discursive stakes to interpretive pragmatic categories of agentive attribution and content framing. Inter-annotator agreement reached at least substantial levels across all categories (κ ≥ .60). Applying the framework to a corpus of periphrastic passives from a Corporate Social Responsibility (CSR) report, we show that content framing most frequently foregrounds positive company actions, consistent with the use of the periphrastic passive as a potential resource for managing agent visibility to project a more benevolent image of the company.