From form to function: an ontology of communicative influence beyond truth and intent

Abstract The widespread use of AI-generated slopaganda and influence campaigns creates a need for computational analysis at scale. Research on disinformation has largely asked whether messages are true or intended to deceive. Yet much communicative influence operates without falsity, through framing, narrative, group positioning, and orientation toward civic or commercial outcomes. Existing frameworks address complementary aspects of such influence but do not represent them as distinct dimensions of the same message; this paper presents an ontology that does. The ontology is truth- and intent-agnostic and operates at the level of message form, resting on the premise that strategic communication is goal-directed and tends to leave observable traces in the message itself. It distinguishes four conceptually orthogonal dimensions: Method (how influence operates), Narrative (how meaning is organised), Polarization (how groups are positioned), and AboutObject (what broader outcome the message bears toward), while TopicDomain independently indexes what the message is about. A single inferential primitive, the cue, grounds base-level labels in message evidence, from which higher-level labels are derived through fixed mappings. Discriminant validity is assessed through irreducibility and content validity through systematic framework mapping; formal verification checks logical coherence and structural and derivation constraints. Empirical annotation reliability and computational annotation performance are not assessed. The resulting structured representation supports both close analysis of individual messages and comparison across large corpora. The codebook, formal verification artefacts, and worked annotation examples are released as open supplemental material.