Learning Objectives

  • Recognise causal fallacies and explain why correlation is not causation
  • Identify weak evidence fallacies — hasty generalisation, anecdotal, and Texas sharpshooter
  • Spot circular reasoning and the burden of proof fallacy
  • Recognise linguistic fallacies — ambiguity, appeal to nature, and middle ground
  • Apply the full set of informal fallacies to evaluate real arguments
1

Causal Fallacies — Correlation is Not Causation

Causal fallacies (also called non causa pro causa — "not the cause for the cause") occur when a connection between events is mistaken for a causal relationship. Establishing genuine causation requires more than identifying a correlation.

Assuming that because one event followed another, the first caused the second.

"I started wearing a lucky bracelet, and since then I have not had a cold. The bracelet is preventing illness."

Temporal sequence (A before B) is necessary for causation but not sufficient. Many things happen before many other things without causing them.

Assuming that two things that occur together are causally related. This is the classic correlation-causation confusion.

"Ice cream sales and drowning rates both increase in summer. Ice cream causes drowning."

Both are caused by a third variable (hot weather), not by each other.

Any error in which a non-causal relationship is treated as causal. This is the umbrella term for all causal fallacies.

"Countries with more television sets per household have higher life expectancy. Therefore, buying more televisions increases life expectancy."

Claiming that if one event is allowed to occur, a chain of increasingly bad consequences will inevitably follow — without justifying why each step in the chain is necessary.

"If we allow students to retake one exam, soon they will expect to retake every assessment, and eventually grades will become meaningless."

Post Hoc in action — video

Watch this clip illustrating post hoc ergo propter hoc reasoning in a US television drama.

Post hoc ergo propter hoc — West Wing

2

Weak Evidence Fallacies

These fallacies involve using evidence that is technically present but insufficient to support the conclusion drawn.

Drawing a broad conclusion from a sample that is too small or unrepresentative to support it.

"I met two people from that city who were rude. People from that city are rude."

Using a personal experience or a single isolated case in place of systematic evidence.

"My grandfather smoked every day and lived to 95. Smoking cannot be that dangerous."

Selecting data that supports a predetermined conclusion while ignoring the rest. The name comes from a mythical shooter who fires at a barn, then draws a target around the bullet holes.

"This dietary supplement has been tested 100 times. In 6 trials it outperformed the placebo — proof that it works." (The arguer ignores the 94 trials showing no effect.)

Drawing conclusions from a sample that does not reflect the population being studied.

"Our survey of 500 social media users found that 90% support this policy. The public overwhelmingly supports it." (Social media users are not a representative sample of the general public.)

Using statistics as evidence when they are biased, cherry-picked, or inapplicable to the claim.

"Our new drug reduces the risk of heart attack by 50%." (The risk fell from 2% to 1% — a 50% relative reduction, but only a 1% absolute reduction in real terms.)

3

Circular and Linguistic Fallacies

Including the conclusion in the premises — or assuming what needs to be proved.

"This book must be true because it says so, and what it says is always true."

Claiming that the burden of proof lies with the person disputing a claim, rather than with the person making it. The principle: extraordinary claims require extraordinary evidence; the burden falls on the claimant.

"Prove that this herbal supplement does not cure cancer. Until you can, I will continue to take it."

A question that presupposes a claim which the respondent has not accepted. Any answer to the question appears to confirm the presupposition.

"Have you stopped cheating on your expenses?" (Answering either yes or no implies the person has cheated.)

Using a word or phrase with more than one meaning and shifting between meanings within an argument.

"The sign said 'fine for parking here.' So I parked there." (The sign uses "fine" to mean a penalty; the driver interprets it as permission.)

Arguing that because something is natural, it is therefore good, safe, or preferable — and that artificial things are bad.

"This product is completely natural, so it has no side effects." (Arsenic and botulinum toxin are natural; many life-saving medicines are synthetic.)

Assuming that a compromise between two positions must be correct. Sometimes the truth is not in the middle — one side may simply be right.

"You say climate change is driven by humans; the other side says it is entirely natural. The truth must be somewhere in the middle."

Claiming that because you cannot understand how something works, it is probably not true or impossible.

"I cannot understand how vaccines train the immune system, so I don't believe they work."

Believing that past random events affect the probability of future independent events. Each spin of a roulette wheel is independent of all previous spins.

"The coin has landed heads ten times in a row. It is bound to be tails next time."

4

Spot the Fallacy

Identify the informal fallacy in each passage. Use both Unit 7 and Unit 8 fallacies.

  1. "Organic food is natural, which means it is healthier than conventional food with its artificial pesticides."
  2. "Kaoru started drinking coffee when she joined Professor X's lab. Twenty years later, Kaoru became a world-famous researcher. This proves the positive effect of drinking coffee on academic achievement."
  3. "Prove to me that extraterrestrial life does not exist. Until you can do so, you have no reason to disbelieve it."
  4. "This new drug reduces stroke risk by 40%!" (In fact, the stroke rate fell from 0.5% to 0.3%.)
  5. "The coin has come up heads eight times in a row. I should definitely bet on tails this time."
  6. "The Bible is the word of God because it says so in the Bible, and the Bible is true because it is the word of God."

  1. Appeal to Nature — "natural" does not mean safe or beneficial. Many natural substances are harmful; many synthetic ones are beneficial.
  2. Post Hoc Ergo Propter Hoc — the coffee drinking preceded her success, but that does not establish it as the cause. Many other factors contributed to her career.
  3. Burden of Proof — the burden of proof lies with the person making the claim, not with those who remain sceptical. "You cannot disprove it" is not evidence for it.
  4. Misleading Statistics — a 40% relative reduction sounds dramatic, but the absolute reduction is only 0.2 percentage points (from 0.5% to 0.3%). The relative framing exaggerates the effect size.
  5. Gambler's Fallacy — each coin flip is an independent event. Past results have no influence on future outcomes for a fair coin.
  6. Begging the Question (Circular Reasoning) — the Bible's authority is assumed to prove its own truth, and its truth is assumed to establish its authority. The argument goes in a circle.
5

Check Your Understanding

Post hoc ergo propter hoc is the assumption that:

Correct! Post hoc ergo propter hoc — 'after this, therefore because of this' — mistakes temporal sequence for causation. Because A happened before B does not mean A caused B. Establishing causation requires controlling for alternative explanations and identifying a plausible mechanism.
Not quite — review the material and try again. Post hoc ergo propter hoc — 'after this, therefore because of this' — mistakes temporal sequence for causation. Because A happened before B does not mean A caused B. Establishing causation requires controlling for alternative explanations and identifying a plausible mechanism.

The Texas Sharpshooter fallacy involves:

Correct! The Texas Sharpshooter fires at a barn and then draws a target around the bullet holes — selecting where the evidence lands to match a conclusion rather than testing the conclusion against all available evidence. Cherry-picking data is the essence of this fallacy.
Not quite — review the material and try again. The Texas Sharpshooter fires at a barn and then draws a target around the bullet holes — selecting where the evidence lands to match a conclusion rather than testing the conclusion against all available evidence. Cherry-picking data is the essence of this fallacy.

Begging the question is a fallacy in which:

Correct! Begging the question (circular reasoning) assumes the truth of the very conclusion it is trying to prove. The premises include the conclusion, either directly or in a slightly disguised form. The argument appears to give a reason, but it is circular — it arrives back where it started.
Not quite — review the material and try again. Begging the question (circular reasoning) assumes the truth of the very conclusion it is trying to prove. The premises include the conclusion, either directly or in a slightly disguised form. The argument appears to give a reason, but it is circular — it arrives back where it started.

Review

Expand each category to consolidate your understanding of Unit 8 fallacies.

Post hoc: A before B, therefore A caused B.
Cum hoc: A and B together, therefore A caused B (or vice versa).
False cause: general causal error — treating correlation as causation.
Slippery slope: claiming that A will lead inevitably to extreme Z without justifying the chain.

Hasty generalisation: too small or unrepresentative a sample.
Anecdotal: a personal experience in place of systematic evidence.
Texas sharpshooter: cherry-picking data to fit a conclusion.
Unrepresentative sample: biased sampling method.
Misleading statistics: inappropriate use of quantitative data.

Begging the question: conclusion embedded in premises.
Burden of proof: shifting responsibility to the sceptic.
Loaded question: question presupposes what needs to be proved.
Ambiguity / equivocation: shifting the meaning of a key term.
Appeal to nature: natural = good.
Middle ground: compromise must be correct.
Personal incredulity: can't understand → must be false.
Gambler's fallacy: past random events affect future independent events.

Key concepts covered in this unit: causal fallacy, post hoc ergo propter hoc, cum hoc ergo propter hoc, false cause, slippery slope, hasty generalisation, anecdotal fallacy, Texas sharpshooter, unrepresentative sample, misleading statistics, begging the question (circular reasoning), burden of proof, loaded question, ambiguity, appeal to nature, middle ground, personal incredulity, gambler's fallacy.

Proceed to Unit 9: Argument Mapping when ready.