Here is a fact you have probably heard many times without ever being told why it is true: a polynomial of degree n has at most n real roots. A cubic can have up to 3. A quartic can have up to 4. A line (n = 1) has exactly one root. The degree is a hard ceiling.
Why exactly n, though? Why not n + 1 or 2n or some other number? The answer is not a coincidence — it is not even a deep theorem that needs calculus. It is a structural fact about how polynomials factor. Every root "costs" you a factor of (x - r), and a polynomial of degree n has only n factors to spend. Once you run out of factors, you run out of roots. That is the whole story.
The factor theorem — the key link
The first piece of the argument is a small but powerful result called the factor theorem.
If r is a root of a polynomial p(x) — that is, if p(r) = 0 — then (x - r) is a factor of p(x). In symbols,
for some polynomial q(x) whose degree is exactly one less than the degree of p. So if p was a cubic, q is a quadratic. If p was a quartic, q is a cubic. The act of peeling off one root shrinks the degree by exactly one.
This is a two-way street. If (x - r) is a factor of p, then setting x = r makes the factor zero, which makes the whole product zero, which means r is a root. Roots and linear factors are the same information, dressed differently.
Each root shrinks the degree by one
Now suppose p(x) has degree n and you have found k distinct roots: r_1, r_2, \dots, r_k. Apply the factor theorem once for each root. After the first root r_1, you can write
with q_1 of degree n - 1. Now r_2 is also a root of p, so plugging it in gives 0. But (r_2 - r_1) is not zero (the roots are distinct), so the zero must come from q_1(r_2) = 0. That makes r_2 a root of q_1, and the factor theorem applies again: q_1(x) = (x - r_2) \cdot q_2(x), with q_2 of degree n - 2.
Keep going. After peeling off all k roots,
Compare degrees. The left side has degree n. The right side has degree k + \deg(q_k), and since \deg(q_k) \geq 0, this is at least k. So
That is the whole proof. A polynomial of degree n has at most n roots because each root costs a factor, and you cannot fit more than n linear factors into a polynomial of degree n without the degree itself growing.
The complex-number version — exactly n roots
Over the real numbers, the bound is \leq n. Over the complex numbers, something stronger holds: a polynomial of degree n \geq 1 has exactly n roots, counted with multiplicity. This is the Fundamental Theorem of Algebra, and it was first proved rigorously by Carl Friedrich Gauss in 1799.
The theorem says the complex numbers are "algebraically closed" — they are large enough that every polynomial equation has all the solutions it is supposed to have. Over the reals, some of those n complex roots may fail to be real, so the count of real roots drops below n. Over the complex numbers, nothing is missing.
So the picture is: a degree-n polynomial always has n roots somewhere — the only question is how many of them are real.
Complex roots come in conjugate pairs
For polynomials with real coefficients (the kind you meet in school), there is a beautiful constraint on the complex roots. If a + bi is a root, then its conjugate a - bi is also a root. The non-real roots always come in matched pairs.
The reason is simple. When you plug a + bi into a polynomial with real coefficients, the imaginary parts come from odd powers of bi. Taking the complex conjugate of the entire equation flips the sign of every i, which is the same as replacing a + bi with a - bi. Since real coefficients are unchanged by conjugation, you get the same equation, and the conjugate is also a root.
One consequence: the number of real roots has the same parity as n. An odd-degree real polynomial must have at least one real root (because complex roots cancel in pairs and leave one unpaired). That is why every cubic graph must cross the x-axis at least once.
Worked example — a cubic with one real root
Consider p(x) = x^3 - 3x^2 + 4x - 2. By inspection, p(1) = 1 - 3 + 4 - 2 = 0, so x = 1 is a root.
Factor it out using polynomial division:
Now the quadratic x^2 - 2x + 2 has discriminant (-2)^2 - 4(1)(2) = 4 - 8 = -4 < 0. No real roots. The complex roots come from the quadratic formula: x = \frac{2 \pm \sqrt{-4}}{2} = 1 \pm i.
So the total root list is \{1, \, 1 + i, \, 1 - i\} — exactly three roots, as the fundamental theorem promises. One is real; the other two are a conjugate pair. The real-root count is 1, which is \leq 3. The bound holds.
Root multiplicity — why "counting with multiplicity" matters
Consider p(x) = (x - 2)^3. This is a degree-3 polynomial, and setting it to zero gives x = 2 as the only solution. Only one root? Not quite — we say x = 2 is a root of multiplicity 3, and it counts as three roots toward the total.
Multiplicity is just the exponent on the linear factor. The polynomial (x - 2)^3 has one factor of (x - 2) repeated three times. When the fundamental theorem says "n roots counting multiplicity," this is what it means: add up the multiplicities and you get n exactly.
The number of distinct real roots can be less than n, but the total factor count never is.
Why the factor theorem is true
The factor theorem is the engine of the whole argument, so it is worth seeing why it is true. Divide p(x) by (x - r) using polynomial long division. The divisor has degree 1, so the remainder has degree 0 — it is a constant. Call it R. Then
This is an identity, true for every x. Evaluate at x = r:
So the remainder equals p(r). (This is called the remainder theorem.) Now the punchline: if r is a root, then p(r) = 0, so R = 0, and the remainder vanishes. The division comes out clean:
That is the factor theorem. Dividing by (x - r) and testing whether r is a root are the same computation.
Why this breaks in modular arithmetic
You might wonder whether "at most n roots" is some universal law. It is not — it depends on the number system. Over the integers modulo 6 (arithmetic where 6 \equiv 0), consider the polynomial p(x) = x^2 - x. Let us check each value:
- p(0) = 0 - 0 = 0.
- p(1) = 1 - 1 = 0.
- p(3) = 9 - 3 = 6 \equiv 0 \pmod 6.
- p(4) = 16 - 4 = 12 \equiv 0 \pmod 6.
Four roots for a degree-2 polynomial. The bound fails.
The reason is that \mathbb{Z}/6\mathbb{Z} has zero divisors — numbers like 2 and 3, neither zero, whose product 6 is zero. The factor theorem secretly relies on the fact that a product (x - r) \cdot q(x) = 0 forces one of the two factors to be zero. In a field (like the reals or the complex numbers), that is always true. In \mathbb{Z}/6\mathbb{Z}, it is not.
School polynomials live over \mathbb{R} or \mathbb{C}, which are fields, so the rule holds. But it is not arithmetic gravity — it is a property of the number system.
Graphical intuition
You can see the bound on the graph. A real root is a place where the curve crosses (or touches) the x-axis. A polynomial of degree n is a smooth curve, and between any two crossings there must be a turning point — a local maximum or minimum. A degree-n polynomial has at most n - 1 turning points (because the derivative has degree n - 1 and therefore at most n - 1 zeros). So the number of crossings is at most one more than the turns: at most n. Geometry and algebra agree.
Quick drill — estimate the root count
- x^2 - 4: degree 2, factors as (x - 2)(x + 2). Two real roots: \pm 2.
- x^2 + 4: degree 2, no real roots; two complex roots \pm 2i.
- x^3 - 1: degree 3, real root 1; two complex roots at the other cube roots of unity.
- x^4 - 1: degree 4, real roots \pm 1, complex roots \pm i. Four total.
- (x - 1)^2: degree 2, one root at 1 with multiplicity 2.
- x^5: degree 5, one root at 0 with multiplicity 5.
In every case, the total root count (with multiplicity, over \mathbb{C}) equals the degree. The real-root count never exceeds it.
The deep reason, in one sentence
The "at most n roots" bound is not a mysterious cap pulled from thin air. Each root you find uses up one linear factor (x - r), and a polynomial of degree n is built from exactly n such factors — no more, no less. Roots are factors in disguise, and the degree is the factor budget.