False Cause

The False Cause fallacy is a family of related fallacies that occur when an arguer gives insufficient evidence for a claim that one thing is the cause of another. All members of this family share the same crime: claiming causation without adequate evidence for it.

graph TD
    FC[False Cause Family]
    FC --> PH[Post Hoc\nErgo Propter Hoc]
    FC --> MC[Mere Correlation]
    FC --> SC[Spurious Correlation]
    FC --> DD[Data Dredging]
    FC --> SS[Slippery Slope]
    FC --> IT[Irrelevant Thesis\nRed Herring]
    PH --> PH1["B followed A\n→ A caused B"]
    MC --> MC1["A and B move together\n→ A caused B"]
    SC --> SC1["Hidden factor C\ncauses both A and B"]
    DD --> DD1["Mine enough data\nand patterns appear by chance"]
    SS --> SS1["A will inevitably\nlead to B, C, D..."]
    IT --> IT1["Divert to unrelated issue\nclaim original is settled"]

How It Appears Per Course

PHIL 252

Covered in Unit 7 as part of the broader study of how scientific and causal claims can fail. The False Cause fallacies extend the taxonomy of informal fallacies from Unit 6 into the domain of causal reasoning and scientific methodology. They are classified as “fallacies of distorting the facts.”

The Six Members

1. Post Hoc Ergo Propter Hoc

“After this, therefore because of it.”

B happened after A, so A must have caused B. The fallacy treats sequence as causation.

Geese arrive in September. Salmon start running later in September. Therefore, the geese cause the salmon run.

The timing is real. The causation is invented. Superstitions are almost always post hoc reasoning.

The test: Could B have happened anyway, without A? If yes — post hoc.

2. Mere Correlation

A and B move together statistically, so A must cause B. But correlation tells you nothing about direction or mechanism.

Countries with higher chocolate consumption tend to win more Nobel Prizes. Therefore chocolate boosts Nobel Prize output.

The correlation may be real. The causation is not established — no mechanism is given, and the direction isn’t argued.

Distinguish from Spurious: Mere correlation = no third factor identified. Spurious = third factor identified.

3. Spurious Correlation

A and C are correlated, but only because a hidden third factor B causes both. Concluding A causes C ignores B entirely.

Ice cream sales and drowning rates are positively correlated. Does ice cream cause drowning?

No — hot weather drives both. The correlation is real and often strong, which is what makes it dangerous.

graph TD
    B[Hidden Factor B\ne.g. hot weather]
    B --> A[Variable A\ne.g. ice cream sales]
    B --> C[Variable C\ne.g. drowning rate]
    A -. "appears to cause" .-> C

4. Data Dredging

Run enough comparisons across enough datasets and something will correlate by pure chance. Data dredging is mining large datasets until a pattern appears, then presenting it as meaningful.

The number of sociology degrees awarded correlates tightly with deaths by anticoagulants.

There is no mechanism. There is no theory. The pattern was found by testing thousands of variable pairs — statistical noise at scale. The more variables you test, the more false positives you will find.

5. Slippery Slope

Claims that A will inevitably lead to B, C, D… without any argument for why each step is inevitable. The fallacy is not that bad outcomes are impossible — it’s that inevitability is assumed, not demonstrated.

“If we allow assisted dying, next we’ll euthanize the disabled, then the elderly, then anyone the state finds inconvenient.”

Each step is asserted. The arguer never shows why step 1 necessitates step 2.

6. Irrelevant Thesis (Red Herring)

Sidestep the original question by raising a compelling but unrelated issue, then claim the original has been answered.

“Should we cut the military budget?” → “Think of the brave soldiers who died for this country!” → Audience feels the question is settled.

The emotional response is real. The argument is a diversion. The original question goes unanswered.

Cross-Course Connections

Causation — all False Cause fallacies involve errors about what causation requires
InformalFallacies — False Cause is a subcategory of informal fallacy
Analogy — False analogy and false cause often co-occur
SelectionBiasVariants — selection bias can produce spurious-looking correlations in data
DataVisualization — charts and graphs are frequently used to present spurious correlations as real

Key Points for Exam/Study

• All False Cause fallacies = claiming causation without adequate evidence
• Post Hoc = sequence only. Mere Correlation = co-movement only. Spurious = hidden third factor.
• Mere vs. Spurious: the difference is whether the hidden third variable has been identified
• Data Dredging = large datasets + no hypothesis + mining for patterns = guaranteed false positives
• Slippery Slope is not always wrong — the fallacy is assuming inevitability without argument
• Irrelevant Thesis / Red Herring: the diversion can be emotionally compelling and still be irrelevant

Open Questions

• Is it ever legitimate to use correlation as evidence for causation, even without a mechanism? What additional conditions would be required?

Cross-course: FalseCause-MotivationTheories — PHIL 252 false cause fallacies applied to evaluating ADMN 201 motivation theory causal claims