Causation

Causation means that one thing produces another — there is a real mechanism connecting them. It is distinct from correlation, which is merely a statistical pattern of two things moving together. Correlation is easy to observe; causation requires more than pattern-matching to establish.

graph TD
    Q[Two variables move together]
    Q --> COR[Correlation\nStatistical pattern only\nNo mechanism claimed]
    Q --> CAU[Causation\nA actually produces B\nMechanism exists]
    COR -->|"mistaking one for the other"| FC[False Cause Fallacy]
    CAU --> P["Probabilistic\n'A raises odds of B'"]
    CAU --> S["Sufficient\n'If A, then always B'"]
    CAU --> N["Necessary\n'Without A, B cannot happen'"]

How It Appears Per Course

PHIL 252

Core topic of Unit 7. The course frames causation as the central target of scientific reasoning, and Unit 7 is concerned with how causal claims can be made legitimately vs. fallaciously. The False Cause family of fallacies all involve mistaking correlation (or coincidence) for causation.

Correlation vs. Causation

Correlation: A and B move together statistically. When A goes up, B goes up (or down). This is measurable and observable but says nothing about why.

Causation: A actually brings about B. Changing A changes B because of a real mechanism.

The danger: correlation is easy to find in large datasets. Causation requires evidence of a mechanism, not just a pattern.

Three Types of Cause

Probabilistic Cause

A raises the chance of B, but doesn’t guarantee it. Most real-world causes are probabilistic.

Smoking increases the probability of lung cancer. Not every smoker gets cancer, but the risk is meaningfully elevated.

Keyword: increases the chance / raises the odds / more likely

Sufficient Cause

If A happens, B always happens. A is enough to guarantee B. But B can also happen without A — A is not the only path to B.

If Sam goes to the party, I go. His going is enough. But I might go for other reasons too.

Keyword: always / guaranteed / enough to produce

Necessary Cause

B cannot happen unless A happens. A is required. But A alone doesn’t guarantee B — it’s a prerequisite, not a guarantee.

Unless Sam goes, I’m not going. Sam is required. But Sam going doesn’t automatically bring me.

Keyword: unless / required / cannot happen without

Quick Reference

Type	If A…	Without A…	Keyword
Probabilistic	B is more likely	B is less likely	raises odds / more likely
Sufficient	B always happens	B might still happen	always / enough
Necessary	B might happen	B cannot happen	unless / required

Cross-Course Connections

FalseCause — the False Cause family all involve errors in causal reasoning
Analogy — scientific hypotheses about causation often begin as analogies
SelectionBiasVariants — a valid causal mechanism can still be undermined by a biased sample
InformalFallacies — False Cause is a category of informal fallacy

Key Points for Exam/Study

• Correlation ≠ causation — never assume causation from pattern alone
• Probabilistic: raises odds (most real causes); Sufficient: always works; Necessary: required
• Sufficient and Necessary are often confused — use the keyword test
• “Always” → Sufficient. “Unless / required” → Necessary. “Raises odds” → Probabilistic
• A cause can be both sufficient AND necessary, but usually they come apart

Open Questions

• Can probabilistic causation be proven, or only inferred statistically? What threshold of probability makes something count as a cause?

Cross-course: Causation-RiskManagement — how causal reasoning underpins the risk management process in ADMN 201 Cross-course: FalseCause-MotivationTheories — PHIL 252 false cause framework applied to motivation theory claims in ADMN 201