🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
division
The fallacy of division is the reverse of composition: it assumes that what is true of the whole must be true of each part. It erroneously distributes properties of an aggregate to its individual members. A wealthy country does not mean every citizen is wealthy; a championship team does not mean every player is a champion caliber performer.
"This university has an excellent reputation, so every professor here must be an excellent teacher."
A job applicant reasons: 'Google is one of the most innovative companies in the world, so every team inside Google must be doing groundbreaking, innovative work.' He's surprised to find the billing department is fairly routine.
A tourist tells friends: 'France has some of the finest cuisine in the world, so that random roadside café I stopped at must serve exceptional food.' She's disappointed by the mediocre sandwich she receives.
Property(X) -> Property(part_of(X))
Binary (yes/no) questions an LLM must answer to identify this aspect:
Does the argument attribute a property of the whole to individual parts?
Type: binaryIs there justification for believing the property applies to each part?
Type: binaryCould the parts have different properties than the whole?
Type: binaryThe fallacy of division is the reverse of composition: it assumes that what is true of the whole must be true of each part. It erroneously distributes properties of an aggregate to its individual members. A wealthy country does not mean every citizen is wealthy; a championship team does not mean every player is a champion caliber performer.
People use group-level information as a shortcut for judging individuals, which works often enough to feel reliable. The error lies in ignoring the variation within groups.
Ask whether the property necessarily applies to every member or is an aggregate/average measure. Highlight variation within the group.
Appears in stereotyping based on national or institutional identity, investment decisions ('this is a great company so all its products must be great'), and education policy where school-level metrics are applied to individual students.
Use these tools to detect, analyze, or train this aspect.