Every math class so far asked “what’s the answer?” Geometry asks something harder: how do you know? Examples suggest. Only a proof settles.
🧠Two Engines of Thought
Inductive reasoning looks at cases and guesses the rule. Deductive reasoning starts from facts everyone already accepts and forces a conclusion. You need both — but only one of them produces certainty.
💡Induction is not second-rate. Almost every idea in mathematics was born by somebody noticing a pattern. The rule is simply: induction proposes, deduction proves. Never let step one wear the badge of step two.
💣Forty Wins, One Loss — the Loss Settles It
Feed whole numbers into n² + n + 41. Every input from 0 to 39 spits out a prime. Then the streak dies with no warning. That crash is the most important thing on this poster.
💻Euler, 1769: no fifth power is the sum of only three or four fifth powers. It stood 197 years. In 1966 Lander and Parkin ran a computer search and published one line: 27⁵ + 84⁵ + 110⁵ + 133⁵ = 144⁵ (both sides are 61,917,364,224). That was the whole refutation — a counterexample doesn’t have to argue, it only has to exist.
🔁Take an If-Then Apart
A conditional is “If p, then q” — p is the hypothesis, q is the conclusion. It has three relatives, and they do not all behave the same way. Test them on: “If a figure is a square, then it is a rectangle.”
Name
Form
Tested on the square example
Conditional
If p, then q
“square → rectangle” TRUE
Converse
If q, then p
“rectangle → square” FALSE — a 2 cm × 5 cm rectangle kills it
Inverse
If not p, then not q
“not square → not rectangle” FALSE — same 2 × 5 counterexample
Contrapositive
If not q, then not p
“not rectangle → not square” TRUE — always matches the original
✅Contrapositive — a safe swap
Original
If it is a dog, then it is a mammal.
Contrapositive
If it is not a mammal, then it is not a dog.
Same sentence viewed from the other side. Truth value is guaranteed identical.
❌Converse — an illegal swap
Original
If it is a dog, then it is a mammal.
Converse
If it is a mammal, then it is a dog.
A cat. The converse is a brand-new claim that must be tested from scratch.
⚠️Confusing a statement with its converse is the most common reasoning error people make — in geometry and everywhere else. “All thieves leave fingerprints” does not make the person with fingerprints a thief. Check the direction of the arrow every single time.
🏗️Every Proof Stands on Something
You cannot justify each statement with an earlier one forever — eventually the chain must rest on bedrock. Around 300 BCE Euclid solved this: write down a short list of statements accepted without proof, then derive everything else.
🧩Postulate vs TheoremA postulate is accepted without proof. A theorem has been proved from postulates, definitions, and earlier theorems. Both are legal reasons inside a proof — the difference is how they earned the right to be there.
✏️Your First Theorem: Vertical Angles
Two lines cross and make four angles. The two opposite wedges are vertical angles. Everyone can see they look equal. Here is how you know it — for every pair of intersecting lines that has ever been drawn.
THEOREM · Vertical angles are congruent
G
GIVEN: two lines intersect, forming ∠1, ∠2, ∠3, ∠4 in order around the point. PROVE:∠1 ≅ ∠3
#
Statements
Reasons
1
Two lines intersect, forming ∠1, ∠2, ∠3, ∠4
Given
2
m∠1 + m∠2 = 180° m∠2 + m∠3 = 180°
Linear Pair Postulate — a linear pair is supplementary
3
m∠1 + m∠2 = m∠2 + m∠3
Substitution — both sums equal 180°
4
m∠1 = m∠3
Subtraction Property of Equality — remove m∠2 from both sides
5
∠1 ≅ ∠3
Definition of congruent angles (equal measures)
⭐ Five rows. Five reasons. No gaps. — and it covers every pair of crossing lines at once.
!
The rule of the two-column proof: no reason, no row. Read the left column and you get the claim; read the right column and you get the permission slip for every line. If you cannot name the reason, the step is not allowed — no matter how obvious the picture looks.
🔧Cash the Theorem In
Once vertical angles are proved congruent, that theorem becomes a legal reason inside every future problem. Here it turns a picture into an equation.
Two lines cross. One angle is (5x + 10)°; the angle vertical to it is (3x + 40)°. Find both.
1
Name the relationship. They are vertical angles, so by the theorem they are congruent — which means their measures are equal.
2
Write the equation the reason demands: 5x + 10 = 3x + 40
Substitute back — always. 5(15) + 10 = 75 + 10 = 85 and 3(15) + 40 = 45 + 40 = 85. They agree, so the setup was right.
5
Extend if asked. The angle next to either one is its linear-pair partner: 180 − 85 = 95°. All four angles are 85°, 95°, 85°, 95°.
⭐ x = 15, and each vertical angle measures 85°
🔑Law of Detachment. If “If p, then q” is true and p is true, then q is true. You know “if a quadrilateral is a rhombus, all four sides are congruent,” and you are told ABCD is a rhombus — so you may detach the conclusion. That single move is what turns a stack of theorems into a working toolbox.
🔑Key Terms
📈Inductive ReasoningSpecific cases → a general claim. Produces ideas worth testing; the conclusion is only probable, no matter how many cases you checked.
🔗Deductive ReasoningAccepted facts + valid logic → a conclusion that must be true if the starting facts are. This is what a proof is made of.
💭ConjectureAn unproved statement you believe, usually from a pattern. A claim on trial: it stays a conjecture until proved, or until a counterexample kills it.
💣CounterexampleOne case that meets the hypothesis and fails the conclusion. A single valid one proves a general statement false, permanently.
➡️Conditional Statement“If p, then q.” The part after if is the hypothesis; the part after then is the conclusion.
🔄ConverseSwap hypothesis and conclusion: “If q, then p.” A true conditional can have a completely false converse — the swap is not truth-preserving.
⚖️ContrapositiveNegate both parts and swap: “If not q, then not p.” Always logically equivalent to the original statement.
🪹Postulate (Axiom)A basic statement accepted as true without proof. The agreed foundation every proof is allowed to build on.
🎖TheoremA statement proved from definitions, postulates, and earlier theorems. Once proved, it becomes a legal reason inside future proofs.
📋Two-Column ProofStatements on the left, the reason justifying each on the right. The governing rule: no reason, no row.
🌎Where This Shows Up
⚖️Courtrooms and crime labs. Forensic scientists reason inductively — “these fibres match, so they probably share a source” — while lawyers push toward deductive certainty. That gap is exactly why expert testimony is phrased as “consistent with” rather than “proves”. Telling a strong pattern apart from a forced conclusion is a legal skill before it is a math skill.
✈️Software that cannot be allowed to fail. Flight-control code, pacemaker firmware, and cryptographic protocols get checked by formal verification: machine-checked proofs that the program behaves correctly for every possible input. Testing a million inputs is induction — and can still miss the one that crashes the plane. A proof covers the untested cases too.
📜A 350-year cliffhanger. Around 1637 Pierre de Fermat scribbled a claim in a book margin, saying he had a marvellous proof the margin was too small to hold. Mathematicians verified it across enormous ranges of numbers — which proved nothing. Only when Andrew Wiles completed a genuine proof in the mid-1990s did it stop being Fermat’s conjecture and become Fermat’s Last Theorem.
📌Remember This
1Induction generates conjectures; deduction certifies them. Both are essential, but only deduction produces certainty. Examples suggest — proofs settle.
2A general claim needs a proof, but it takes only one counterexample to destroy it. That asymmetry is why mathematicians hunt counterexamples first and write proofs second.
3A conditional and its contrapositive always share a truth value; a conditional and its converse do not. Swapping “if” and “then” is not a legal move — check it every time.
🤔 Think about it
You test a conjecture on the first one billion whole numbers and it holds every single time. Would you bet your savings on it? Would a mathematician? Name exactly what separates “overwhelming evidence” from “proof” — and why mathematics refuses to blur a line that almost every other field is happy to blur.
Every proof starts from postulates — statements accepted without proof. So is a theorem really “true”, or only “true given these assumptions”? When mathematicians changed one of Euclid’s postulates in the 1800s they got perfectly consistent non-Euclidean geometries — and one of them later turned out to describe gravity in Einstein’s universe.
⭐Remember: a pattern is a question, not an answer. Say what you assumed, name the reason for every step, and hunt for the one case that breaks it — before someone else does.