🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
faulty_agency_assignment
Faulty agency assignment occurs when responsibility for an outcome is attributed to the wrong actor, force, or factor. This includes blaming victims for systemic failures, crediting leaders for trends they did not influence, attributing outcomes to deliberate intent when they resulted from accident or structural forces, or diffusing responsibility across so many actors that no one is held accountable. It is a discourse mechanic that shapes narratives of credit and blame.
After a building collapse that killed dozens, officials blame the residents for 'ignoring evacuation warnings' rather than addressing the building code violations, corrupt inspectors, and negligent developers who allowed an unsafe structure to be occupied in the first place.
After a data breach exposes millions of customers' personal information due to a company's failure to update its security software, the company's PR team releases a statement urging customers to 'be more vigilant about monitoring their accounts and using strong passwords,' deflecting from the company's own negligence.
A local politician, responding to a surge in pedestrian accidents at an intersection with a known design flaw, says: 'People need to put their phones away and pay attention when they cross the street.' No mention is made of the missing crosswalk signals, poor lighting, or repeated requests from residents for safety improvements.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Does the text use collective pronouns ('We must...', 'Humanity should...')?
Type: binaryIs responsibility or guilt assigned to a massive, diffuse group?
Type: binaryDoes the group lack the cohesive agency to act as described?
Type: binaryFaulty agency assignment occurs when responsibility for an outcome is attributed to the wrong actor, force, or factor. This includes blaming victims for systemic failures, crediting leaders for trends they did not influence, attributing outcomes to deliberate intent when they resulted from accident or structural forces, or diffusing responsibility across so many actors that no one is held accountable. It is a discourse mechanic that shapes narratives of credit and blame.
Assigning agency to a specific actor creates a simple, satisfying narrative with a clear protagonist or villain. It is cognitively easier to blame a person than to understand a complex system, and it protects powerful actors when blame is displaced onto the powerless.
Map the full causal chain and identify all actors who contributed to the outcome. Ask who had the most power to prevent the negative outcome and whether the attributed agent actually had the capacity and information to act differently.
Faulty agency assignment appears in corporate crisis communications (blaming workers for system failures), political scapegoating, environmental disasters (blaming individual consumers instead of industrial polluters), and financial crises.
Drawing broad conclusions from limited, unrepresentative, or anecdotal evidence.
Proposing individual solutions to problems requiring systemic/collective enforcement.
Anthropomorphisation as a fallacy occurs when human characteristics such as desires, intentions, beliefs, or emotions are attributed to non-human entities — animals, algorithms, corporations, natural phenomena — and these attributed qualities are then used as the basis for reasoning or argumentation. While anthropomorphic language can be a useful heuristic, it becomes fallacious when the projected human qualities are treated as literal truths that drive conclusions.
The mereological fallacy involves a confusion between the properties of parts and the properties of wholes, but differs from the simpler composition and division fallacies in that it involves a category error about what kind of entity can possess a given property. While composition/division involve incorrect inferences about the same type of property at different levels, the mereological fallacy attributes properties to entities at a level where those properties are conceptually inapplicable — as when neuroscientists say 'the brain decides' or 'the hippocampus remembers,' attributing person-level psychological predicates to sub-personal components.
Use these tools to detect, analyze, or train this aspect.