Why the distinctions matter
Different reasoning failures call for different questions, different teaching moves, and different repairs. A causal blunder is not corrected the same way as a formal invalidity or a sampling mistake.
Theory article
Formal, Informal, Causal, and Statistical Kinds of Reasoning Failure
People often talk as if a logical fallacy were a single species of badness. It is not. Some mistakes are structural, some are evidential, some are causal, and some are statistical. Throwing them all into one bucket is a bit like teaching students that bones, blood sugar, and broken eyeglasses are all just 'health problems' and then acting surprised when treatment becomes vague.
Traditional and site-specific taxonomies
Traditional logic often contrasts formal and informal fallacies. LogFall keeps that insight but also uses more teaching-friendly families such as evidential, causal, statistical, linguistic, and relevance-based failures.
Four large kinds of failure
These are not the only categories, but they are the most pedagogically useful first cuts.
Formal or structural
The support fails because the conclusion does not follow from the shape of the argument. Cases like Affirming the consequent and Denying the antecedent belong here.
Evidential
The speaker misuses, selects, or overreads the evidence. Cherry picking and Absence of evidence fallacy are classic cases.
Causal or explanatory
The problem concerns what caused what, what mechanism is missing, or what explanation is too thin. Post hoc ergo propter hoc lives here, as does Wrong causal direction.
Statistical
The failure lies in rates, samples, distributions, uncertainty, or comparison classes. Base rate fallacy and Survivorship bias are good examples.
What each kind needs from the teacher
A good class changes the tool to match the failure.
Formal errors need structure made visible
Students need short premises, short conclusions, and often a map or symbolic skeleton. Long political prose is a terrible first home for teaching invalid form.
Evidential errors need comparison sets
Ask what evidence is missing, selected, exaggerated, or treated as sufficient. The cure here is often not formal notation but a fuller record.
Causal errors need alternative explanations
Students should be trained to ask about sequence, mechanism, reverse causation, third variables, and counterfactual alternatives.
Statistical errors need numerical humility
Base rates, sample size, variance, regression, and uncertainty language usually matter more here than rhetoric does. The math does not have to be fancy; it just has to be present.
Common category mistakes
Misclassification produces weak teaching and weak criticism.
Treating every informal mistake as 'emotional'
Some arguments are vivid or heated, but their deepest problem may be sampling, category confusion, or causal overreach rather than emotional pressure.
Treating every causal mistake as a formal mistake
A causal argument may be invalid in some abstract rendering, but that often hides the more teachable point: the mechanism or evidence is inadequate.
Forgetting that one passage can host multiple kinds
A speech can mix Ad hominem with Cherry picking and a statistical overreach in the same paragraph. Bad reasoning is perfectly capable of multitasking.
Using family labels as if they were verdicts
A family label should point students toward the right diagnostic questions. It should not replace the more specific diagnosis.
Takeaway
Different reasoning failures need different lenses.
If you teach every bad argument with the same voice and the same tool, students will learn the vocabulary while missing the craft. The categories matter because the remedies matter.
References and further reading
Sources that ground the article or push the discussion further.
Fallacies (Stanford Encyclopedia of Philosophy)
Fallacies (Stanford Encyclopedia of Philosophy) — Strong background on formal and informal distinctions and on modern fallacy theory.
Informal Fallacies (OpenStax Introduction to Philosophy)
Informal Fallacies (OpenStax Introduction to Philosophy) — Useful teaching taxonomy for relevance, weak induction, unwarranted assumption, and diversion.
Inductive Logic (Stanford Encyclopedia of Philosophy)
Inductive Logic (Stanford Encyclopedia of Philosophy) — Helpful on strong and weak inductive support, especially where probability enters the picture.
Philosophy of Statistics (Stanford Encyclopedia of Philosophy)
Philosophy of Statistics (Stanford Encyclopedia of Philosophy) — Useful for the statistical side of evidence, inference, and methodological reasoning.