Passive voice functions as a core diffusion layer in institutional language systems. It evaporates explicit agency, converts directed actions into ambient processes, and manufactures victimless events at scale.1
The Diffusion Layer
In active voice, causal chains remain intact: "The executive team approved the layoffs." The subject carries load. In passive voice, the load is distributed into the ether: "Layoffs were approved."
The transformation does not merely obscure the actor — it re-engineers the ontological status of the event itself. Responsibility becomes a floating nominalization.2
Causal Chain Visualizer
Sentence Construction Converter
Test the diffusion engine. Input active voice → observe full agency evaporation.
Case Stack
The passive construction is not automatically dishonest. Sometimes the actor is unknown, irrelevant, or deliberately withheld for ethical reasons. But in institutional speech, the pattern becomes diagnostic when the actor is obvious, powerful, and missing.
The board closed the plant.
The plant was closed.
Police fired tear gas into the crowd.
Tear gas was deployed.
The platform changed the ranking rules.
Ranking rules were updated.
Case studies. The paired sentences above are illustrative constructions, built to isolate the transformation — not quotations attributed to named actors. The pattern they model, however, is documented across the critical-linguistics literature on agentless passives in institutional and news discourse.1
System Effects
Actor Deletion
The sentence preserves the injury while removing the hand that produced it.
Process Naturalization
Decisions begin to resemble weather: regrettable, external, and nobody's direct responsibility.3
Response Management
Public attention is redirected from accountability toward adaptation, mitigation, and emotional closure.
Restoration Protocol
To reverse agency laundering, do not merely ask whether a sentence is grammatically passive. Ask what work the passivity is performing.
- Locate the event: what happened?
- Locate the affected object: who or what absorbed the consequence?
- Reinsert the actor: who had the authority, incentive, or mechanism to make it happen?
- Restore the verb: replace atmospheric process language with the concrete action.
- Read the sentence again and measure the political difference.
Method
This is a reading, not a measurement. The claim is narrow and falsifiable: that the agentless passive — the passive with its by-phrase deleted — systematically removes the actor from a clause while preserving the event, and that institutions reach for this construction disproportionately when the actor is powerful and the action is costly.
The evidence is the grammar itself. Following the transitivity analysis of systemic-functional linguistics, every clause encodes a choice about who is named as the doer; the agentless passive is the option that names no one.2 The diagnostic procedure is the Restoration Protocol above: take a suspect sentence, attempt to reinsert a specific actor, and observe whether the meaning that returns was being actively suppressed.
Limits: grammar is evidence of construction, not of intent. That a sentence is agentless does not prove anyone chose it to evade — only that the form is available for evasion, and that its distribution in institutional speech is not random. The method diagnoses a pattern; it does not read minds.
Counter-positions
The strongest objection comes from descriptive grammarians, most pointedly Geoffrey Pullum, who argues that the passive is chronically misdiagnosed and unfairly maligned — that most writing advice cannot even identify a passive correctly, and that the construction has entirely legitimate functions: managing information flow, foregrounding the affected party, and omitting an actor who is genuinely unknown or irrelevant.4
We concede the linguistics entirely. The passive is not a defect, and "avoid passive voice" is bad advice. That is precisely why the construction works as a laundering layer: its ordinariness is the camouflage. Our claim is not that the passive is bad grammar but that the agentless passive, deployed where the actor is obvious and powerful, performs political work — and that the work is invisible exactly because the form is so unremarkable. Pullum shows the tool is neutral; we are describing one of its uses.
Notes & Sources
1 Fowler, Roger, Bob Hodge, Gunther Kress, and Tony Trew. Language and Control. Routledge & Kegan Paul, 1979. (Critical linguistics on syntactic transformations that delete agency.)
2 Halliday, M.A.K. An Introduction to Functional Grammar. Edward Arnold, 1985. (Transitivity and the clause as a configuration of participant roles.)
3 Orwell, George. "Politics and the English Language." Horizon, 1946. (Political language engineered to make the indefensible sound natural and inevitable.)
4 Pullum, Geoffrey K. "Fear and Loathing of the English Passive." Language & Communication 40 (2014): 60–66.
Additional foundations: Fairclough, Norman. Language and Power. Longman, 1989.
Method and apparatus follow the project's Research Standards.