Topic: Logic
Errors in the structure of an argument: inferences whose flaw is visible from the logical form alone, regardless of whether the individual claims happen to be true.
| Fallacy | Trigger – look for… |
|---|---|
| Denying the antecedent | A conditional (“If P then Q”) is misused: because P does not hold, Q is denied – even though Q could have another cause. (Checkable form: “If P → Q; not-P → ∴ not-Q”). |
| Affirming the consequent | A conditional (“If P then Q”) is misused: because Q holds, P is affirmed – even though Q could have another cause. (Checkable form: “If P → Q; Q → ∴ P”). |
| Affirmative conclusion from a negative premise | In a formal categorical inference (especially a syllogism), an affirmative conclusion (“all … are …”, “some … are …”) is inferred although at least one premise is negative (“no …”, “some … are not”). |
| Negative conclusion from affirmative premises | In a formal categorical inference (especially a syllogism), a negative conclusion (“no …”, “some … are not”) is inferred although all premises are affirmative (“all … are …”, “some … are …”). |
| Fallacy of exclusive premises | In a standard categorical syllogism or comparable form, all premises are negative, yet a conclusion is still derived. |
| Undistributed middle | In a formal categorical inference (especially a syllogism), the middle term, which connects the premises, refers only to a part of its class in all occurrences (typically as predicate of “All X are M”, or within an unqualified “Some …” statement), yet is still used as if it can connect the other terms. (Checkable form: “Some C are M; All P are M → ∴ …P, C…”.) |
| Illicit process | In a formal categorical inference (especially a syllogism), the major or minor term occurs in a distributed position (whole-class reference) in the conclusion, yet only in an undistributed position in its premise – i.e., the conclusion makes a broader claim about that class than the premise does. (Checkable form: “All P are M; All M are S → ∴ All S are P” – “S” is broader in the conclusion.) |
| Fallacies of distribution | A categorical inference (especially a syllogism) in which the distribution of terms is evidently mishandled but does not clearly match either specific rule above – e.g., several distribution errors at once, or a non-standard form where the middle/major/minor roles are hard to assign (enthymemes with elided premises, multi-premise chains). Fallback entry: if the conditions for undistributed middle or illicit process are met, use those instead. |
| Ambiguous middle term | In a syllogism, the middle term shifts meaning between premises (equivocation), so it fails to link the other two terms. (A specific form of the four-term fallacy.) |
| Four-term fallacy | A syllogism contains four distinct terms instead of exactly three – often because a term is used in two different meanings (equivocation). Count meanings, not words. |
| Circular reasoning | A conclusion is supported by a premise that directly or indirectly depends on that same conclusion, including through a chain of claims that loops back to it. |
| Existential fallacy | An existential conclusion (“some A are B”) drawn from universal premises (“all A are B”) that do not establish the class actually exists (vacuously true if A is empty). |
| Fallacy of division | A property of the whole is inferred to hold for its parts (or a group-level property for individual members) – as if the parts must share the whole's property, without checking whether it is emergent. |
| Fallacy of composition | A property of the parts (components, members) of a system is inferred to hold for the whole as well – as if the whole must be the sum of its parts, without checking for emergent or resultant properties. |
| Fallacies of emergence | A property is transferred between a whole and its parts (or a group and its members) as if the levels were interchangeable – without checking whether the property emerges upon combination. Umbrella/fallback: route to either fallacy of division or composition if the direction is clear. |
#END.