Fallacies of emergence

Core claim: A property is transferred between a whole (system, group, population, period) and its parts (components, members, instances, instants) as if the two levels were interchangeable – ignoring that properties may emerge (appear or disappear) when parts combine. This is an umbrella entry for two specific fallacies: fallacy of composition (parts → whole) and fallacy of division (whole → parts).

• A property of a whole is attributed to its parts, or a property of the parts is attributed to the whole, without checking whether the property actually holds at the target level.
• The argument treats the whole’s property as the sum (or average) of the parts’ properties – or vice versa – ignoring that interactions between the parts can change, cancel, or produce new properties.
• A property that exists only across a span or period (the whole) is attributed to a single point or instance (a part) – or the reverse: properties of individual instants are assumed to determine the behaviour over the whole period (e.g., “the arrow is at rest at every instant, therefore it never moves”).

• Routing: if the direction of the transfer is clear, use the specific fallacy instead:
  • Parts → whole (properties of the components inferred for the whole) → Fallacy of composition (MD)
  • Whole → parts (properties of the whole inferred for the parts) → Fallacy of division (MD)
• This entry applies when the direction is unclear, when both directions occur in the same argument, or when a whole/parts error is detected but does not cleanly match either specific pattern.
• Underlying phenomenon: emergence – properties that appear in a system or group as a whole that are not recognisably inherent in its parts. The concept also covers the disappearance of properties upon combination (sodium and chlorine are both toxic, but table salt is not). The opposite is resultance: properties of the whole that can be inferred from the properties of the components.
• Statistical counterparts: the same error at the level of aggregation (group average ↔ individual) is the Ecological fallacy (MD) (group → individual) and the atomistic fallacy (individual → group; page pending).
• The distinction between emergent and resultant properties is difficult and debated (especially in physics and chemistry) – but the fact that drawing the line is hard does not imply that the distinction does not exist.

• The property is genuinely additive: the whole’s property is the sum of identical part-level properties (e.g., total mass = sum of the masses), so the transfer is valid in both directions.
• The property is constitutive of the parts by definition (e.g., “the mosaic is made of tiles, so this section of it is made of tiles”).
• The transfer is explicitly framed as an approximation or heuristic, with the uncertainty kept in mind.
• The discussion is about emergence itself (philosophical or scientific analysis of emergent properties) rather than making an inference from one level to the other.

• Fallacy of composition (MD) – parts → whole. Ask: which direction does the transfer go?
• Fallacy of division (MD) – whole → parts. Ask: which direction does the transfer go?
• Ecological fallacy (MD) – the statistical version: a group average is projected onto an individual or subgroup. Ask: is the whole’s property a statistical measure (→ ecological), or a qualitative or structural property (→ this family)?
• Mereological fallacy (MD) – conflates the properties of a system as a whole with those of its functional components. Ask: is it a directional transfer of a specific property (→ this family), or a general conflation of the two levels (→ mereological)?
• Atomistic fallacy (MD) – individual → group: extrapolating from individuals or small groups to a larger population, ignoring emergent properties. Ask: is the transfer statistical (→ atomistic / ecological), or qualitative (→ this family)?

Input: “The meadow is multicoloured, so this flower growing on it must be multicoloured too.”
Output: Whole → parts: “multicolouredness” only emerges from many differently coloured flowers occurring together in one place. Route to Fallacy of division (MD).

Input: “Every player on this team is the best at their position, so this must be the best team.”
Output: Parts → whole: “being the best team” is an emergent property of the players’ interplay, not the sum of their individual qualities. Route to Fallacy of composition (MD).

Input: “The average income in this country is 50,000, so a typical resident must earn about that.”
Output: This is a statistical whole/parts transfer – a group average projected onto an individual. Route to Ecological fallacy (MD).

Input (legitimate): “The total weight of the shipment is the sum of the weights of the individual packages.”
Output: Mass is an additive property, so the whole/parts transfer is valid here. Not a fallacy.

Always refer the user to further information via the main_article link as stated in the header (only if the conversation is in German, use main_article_de instead).