Logic

The formal structure of reasoning: how to determine whether conclusions actually follow from premises.

Why Logic Matters

Logic is the scaffolding of all reasoning. Without it, you can't tell whether a conclusion follows from its premises. You're just guessing.

You don't need to be a logician. But understanding the basics lets you:

  • Spot when someone's reasoning is broken regardless of how persuasive they sound
  • Construct arguments that actually hold up
  • Distinguish between "this feels right" and "this is right"

Three Types of Reasoning

TypeDirectionStrengthExample
DeductiveGeneral → SpecificCertain (if valid)All mammals breathe air. Dogs are mammals. Therefore dogs breathe air.
InductiveSpecific → GeneralProbableEvery dog I've seen breathes air. Therefore all dogs probably breathe air.
AbductiveObservation → Best ExplanationPlausibleThe grass is wet. The best explanation is that it rained.

Deductive Reasoning

Starts with general principles and derives specific conclusions. If the premises are true and the logic is valid, the conclusion must be true.

Structure:

Premise 1: All A are B
Premise 2: C is A
Conclusion: Therefore, C is B

Strength: Guarantees truth (if premises are true and form is valid). Weakness: Can only produce conclusions already contained in the premises. Only as good as your starting premises.

Everyday examples:

  • "The store closes at 9. It's 9:30. Therefore the store is closed."
  • "You need a passport to travel internationally. I don't have a passport. Therefore I can't travel internationally."

Inductive Reasoning

Starts with specific observations and draws general conclusions. The conclusion goes beyond what the premises guarantee.

Structure:

Observation 1: Swan 1 is white
Observation 2: Swan 2 is white
...
Observation 1000: Swan 1000 is white
Conclusion: All swans are probably white

Strength: Generates new knowledge; the basis of scientific discovery. Weakness: Never 100% certain. One black swan disproves the conclusion.

Strong vs. Weak Induction:

Strong InductionWeak Induction
Large, representative sampleSmall or biased sample
Consistent observationsMixed results
No known counterexamplesKnown exceptions ignored
Conclusion is modestConclusion overstates the evidence

Abductive Reasoning

Starts with an observation and infers the most likely explanation. Also called "inference to the best explanation."

Structure:

Observation: The grass is wet
Possible explanations: Rain, sprinklers, dew, flooding
Best explanation: It rained (given the time, weather, scope)
Conclusion: It probably rained

Strength: How we actually reason most of the time. Essential for diagnosis, detective work, and everyday problem-solving. Weakness: "Best available explanation" isn't the same as "correct explanation." Better explanations may emerge.

Criteria for the "best" explanation:

CriterionMeaning
SimplicityFewer assumptions (Occam's Razor)
Explanatory scopeAccounts for more observations
ConsistencyFits with existing knowledge
TestabilityMakes predictions that can be checked
FruitfulnessLeads to new insights or predictions

Validity and Soundness

Two critical concepts that most people confuse.

ConceptDefinitionRequires True Premises?
ValidThe conclusion follows logically from the premisesNo
SoundValid AND all premises are trueYes

Valid but NOT Sound

Premise 1: All fish can fly        (FALSE)
Premise 2: A salmon is a fish      (TRUE)
Conclusion: A salmon can fly       (LOGICALLY FOLLOWS, but false)

The logic is perfect. The conclusion follows. But the first premise is false, so the argument is unsound.

Sound

Premise 1: All humans are mortal   (TRUE)
Premise 2: Socrates is a human     (TRUE)
Conclusion: Socrates is mortal     (TRUE and logically follows)

Invalid (Even with True Premises)

Premise 1: Some birds can fly      (TRUE)
Premise 2: Penguins are birds      (TRUE)
Conclusion: Penguins can fly       (DOESN'T FOLLOW)

Key insight: An argument can have true premises and a true conclusion and STILL be invalid, if the conclusion doesn't actually follow from the premises.

Common Argument Structures

Modus Ponens (Affirming the Antecedent)

The most fundamental valid argument form.

If P, then Q        (If it rains, the ground gets wet)
P                   (It's raining)
Therefore, Q        (Therefore, the ground is wet)

Valid. Always.

Modus Tollens (Denying the Consequent)

If P, then Q        (If it rains, the ground gets wet)
Not Q               (The ground is not wet)
Therefore, not P    (Therefore, it didn't rain)

Valid. Always. This is the basis of falsification in science.

Hypothetical Syllogism (Chain Reasoning)

If P, then Q        (If I study, I'll pass)
If Q, then R        (If I pass, I'll graduate)
Therefore, if P, then R    (If I study, I'll graduate)

Valid. You can chain as many links as needed.

Disjunctive Syllogism (Process of Elimination)

P or Q              (Either the battery is dead or the starter is broken)
Not P               (The battery is not dead)
Therefore, Q        (Therefore, the starter is broken)

Valid. But only when the "or" is genuinely exhaustive, meaning there aren't other options.

Common INVALID Forms

NameStructureWhy It's Invalid
Affirming the consequentIf P then Q; Q; therefore POther things could cause Q
Denying the antecedentIf P then Q; not P; therefore not QQ might still be true for other reasons

Affirming the consequent example:

If it rains, the ground gets wet.
The ground is wet.
Therefore, it rained.          ← INVALID (sprinklers could have caused it)

Denying the antecedent example:

If it rains, the ground gets wet.
It didn't rain.
Therefore, the ground isn't wet.  ← INVALID (sprinklers exist)

Truth Tables

Truth tables show all possible combinations of truth values for logical operations.

Basic Operators

PQP AND QP OR QNOT PIF P THEN Q
TTTTFT
TFFTFF
FTFTTT
FFFFTT

Key Insights from Truth Tables

"If P then Q" is only false when P is true and Q is false. This confuses people. "If it rains, I'll bring an umbrella" is not violated by bringing an umbrella on a sunny day.

StatementWhat Makes It FALSE
P AND QEither P or Q is false
P OR QBoth P and Q are false
IF P THEN QP is true but Q is false
P IF AND ONLY IF QP and Q have different truth values

Necessary vs. Sufficient Conditions

Condition TypeDefinitionExample
NecessaryMust be true for the conclusion (but alone isn't enough)Oxygen is necessary for fire
SufficientIf true, guarantees the conclusionBeing a dog is sufficient for being a mammal
Necessary AND sufficientThe exact condition, no more, no lessBeing a bachelor is necessary and sufficient for being an unmarried adult male

How to Tell Them Apart

TestQuestionExample
Necessary"Can the conclusion be true WITHOUT this?" If no → necessaryCan you have fire without oxygen? No → oxygen is necessary
Sufficient"Does this ALONE guarantee the conclusion?" If yes → sufficientDoes being a triangle guarantee having 3 sides? Yes → sufficient

Practical Importance

Confusing necessary and sufficient conditions is one of the most common reasoning errors:

ConfusionExampleError
Treating necessary as sufficient"Hard work is necessary for success, so if I work hard I'll succeed"Hard work is necessary but not sufficient. Luck, talent, timing matter too
Treating sufficient as necessary"A degree gets you a good job, so you need a degree for a good job"A degree is sufficient but not necessary. Other paths exist

Logical Equivalences

Some statements say the same thing in different ways.

StatementLogically Equivalent To
If P then QIf not Q then not P (contrapositive)
Not (P and Q)Not P or not Q (De Morgan's)
Not (P or Q)Not P and not Q (De Morgan's)

The contrapositive is always valid:

  • "If it rains, the ground gets wet" = "If the ground isn't wet, it didn't rain"

The converse is NOT equivalent:

  • "If it rains, the ground gets wet" ≠ "If the ground is wet, it rained"

The inverse is NOT equivalent:

  • "If it rains, the ground gets wet" ≠ "If it doesn't rain, the ground isn't wet"

Quantifiers

Words like "all," "some," "none," and "most" change the logic of statements dramatically.

QuantifierTo Disprove, You NeedExample
All X are YOne X that isn't Y"All politicians lie" → One honest politician disproves this
No X are YOne X that is Y"No exercise is fun" → One fun exercise disproves this
Some X are YShow that no X is Y"Some dogs are aggressive" → Prove zero dogs are aggressive
Most X are YShow that half or fewer X are Y"Most startups fail" → Show 50%+ succeed

Common error: Treating "some" as "all" or "most."

"Some studies show X" does NOT mean "studies show X." It means at least one study showed X, which might be an outlier.

Putting It Together: Evaluating Real Arguments

Step-by-Step Process

  1. Identify the conclusion. What's being claimed?
  2. Identify the premises. What reasons are given?
  3. Check the form. Is the argument structure valid?
  4. Check the premises. Are the premises actually true?
  5. Check for hidden premises. What's assumed but not stated?
  6. Assess strength. How strong is the overall argument?

Worked Example

Argument: "Renewable energy is more expensive than fossil fuels. We shouldn't switch to things that cost more. Therefore we shouldn't switch to renewable energy."

StepAnalysis
ConclusionWe shouldn't switch to renewable energy
Premise 1Renewable energy is more expensive than fossil fuels
Premise 2We shouldn't switch to things that cost more
Valid form?Yes (modus ponens structure)
Premise 1 true?Increasingly false: renewables are now cheaper in many cases
Premise 2 true?Too simplistic: ignores externalities, long-term costs, other values
Hidden premisesCost is the only relevant factor; current costs are permanent
AssessmentWeak argument: both premises are questionable and important factors are ignored

Key Takeaways

  1. Validity ≠ Truth. An argument can be logically perfect with false premises
  2. Soundness is what matters. Valid logic + true premises = reliable conclusion
  3. Most reasoning is inductive or abductive. Deductive certainty is rare outside math
  4. Learn the common forms. Modus ponens, modus tollens, and the common invalid forms cover most practical cases
  5. Necessary ≠ sufficient. Confusing these causes enormous errors in everyday reasoning
  6. The contrapositive is your friend. It's always logically equivalent and often more intuitive
  7. Watch the quantifiers. "Some," "most," and "all" are very different claims