Polynomials and Polynomial Arithmetic

A polynomial is a finite sum of terms, each a constant times a non-negative integer power of a variable, e.g. $3 x^{2} - 5 x + 7$ . They can be added, subtracted, multiplied, and divided (with remainder) much like integers. Polynomials are foundational across algebra, calculus, computer science, and physics — used by mathematicians since at least the 9th century (al-Khwārizmī) and formalized over the centuries that followed.

A polynomial in one variable is an expression of the form $a_{n} x^{n} + a_{n - 1} x^{n - 1} + \dots + a_{1} x + a_{0}$ , where the coefficients $a_{i}$ come from a number system (integers, rationals, reals, or complex numbers) and the exponents are non-negative integers. They matter because they are the simplest functions beyond linear ones, yet rich enough to approximate almost any smooth behavior, model curves, and encode data. Polynomial arithmetic — adding, multiplying, and dividing them — mirrors integer arithmetic so closely that the two share a deep structural kinship (both form rings with a division algorithm). A surprising fact: every single-variable polynomial of degree $n$ has exactly $n$ roots when counted with multiplicity in the complex numbers — no more, no fewer — a result so important it is called the Fundamental Theorem of Algebra.

The Core Insight

A polynomial is completely determined by its finite list of coefficients. Therefore all polynomial arithmetic is bookkeeping on coefficient lists: addition combines coefficients of like powers, multiplication is a "convolution" of the two lists, and division produces a quotient and a remainder exactly as long division does for whole numbers. Treating polynomials as objects (not just functions to be evaluated) unlocks factoring, root-finding, and the entire algebraic machinery built on them.

Formal Statement

Polynomial in One Variable

A polynomial in the variable $x$ over a coefficient set $F$ (the field of coefficients, such as $Q$ , $R$ , or $C$ ) is an expression

p (x) = a_{n} x^{n} + a_{n - 1} x^{n - 1} + \dots + a_{1} x + a_{0} = k = 0 \sum n a_{k} x^{k}

⟳This is the definition itself — sigma notation just abbreviates the finite sum, with

x^{0} = 1

a_{0}

is the constant term.

Where the symbols mean:

$x$ — the variable (or indeterminate), a placeholder.
$a_{k} \in F$ — the coefficients; $a_{k}$ multiplies $x^{k}$ .
$a_{0}$ — the constant term.
$n$ — the degree of $p$ , written $de g p$ , provided the leading coefficient $a_{n} \neq = 0$ .
$a_{n} x^{n}$ — the leading term; if $a_{n} = 1$ the polynomial is monic.

The zero polynomial $p (x) = 0$ has, by convention, degree $- \infty$ (or "undefined"). A polynomial of degree 0 is a nonzero constant; degree 1 is linear; degree 2 quadratic; degree 3 cubic; degree 4 quartic; degree 5 quintic.

The Four Arithmetic Operations

Given $p (x) = \sum_{k} a_{k} x^{k}$ and $q (x) = \sum_{k} b_{k} x^{k}$ :

Addition / Subtraction (combine like powers):

(p \pm q) (x) = k \sum (a_{k} \pm b_{k}) x^{k}

⟳Group terms with the same exponent — this is just the distributive and commutative laws applied to a finite sum.

Multiplication (each term times each term, then collect — a convolution):

(p \cdot q) (x) = m = 0 \sum d e g p + d e g q c_{m} x^{m}, c_{m} = i + j = m \sum a_{i} b_{j}

⟳Expand

(\sum a_{i} x^{i}) (\sum b_{j} x^{j})

by distributivity; the coefficient of

x^{m}

collects every product

a_{i} b_{j}

where

i + j = m

Degree Rules

For nonzero polynomials over a field,

de g (p \cdot q) = de g p + de g q, de g (p \pm q) \leq max (de g p, de g q)

⟳The top power of the product is

a_{n} b_{m} x^{n + m}

with

a_{n} b_{m} \neq = 0

since a field has no zero divisors; addition cannot raise the top power but may cancel it.

Division Algorithm for Polynomials

For any polynomials $p (x)$ (the dividend) and $d (x) \neq = 0$ (the divisor) over a field, there exist unique polynomials $q (x)$ (quotient) and $r (x)$ (remainder) with

p (x) = d (x) q (x) + r (x), r (x) = 0 or de g r < de g d

⟳Repeatedly subtract a multiple of

d

that cancels the current leading term of

p

; the degree strictly drops each step, so the process terminates. Uniqueness follows because two valid remainders would differ by a multiple of

d

yet have degree below

de g d

Remainder and Factor Theorems

For a polynomial $p (x)$ and a constant $c$ :

p (x) = (x - c) q (x) + p (c)

so the remainder upon dividing by $(x - c)$ is just $p (c)$ . Consequently $(x - c)$ is a factor of $p$ if and only if $p (c) = 0$ (i.e. $c$ is a root).

⟳Apply [5] with

d (x) = x - c

; the remainder

r

has degree

< 1

, so it is a constant

r

. Substituting

x = c

gives

p (c) = 0 \cdot q (c) + r

, hence

r = p (c)

Why It Holds / Derivation

Polynomial Long Division — Existence of Quotient and Remainder

We prove [5] constructively for $p (x)$ of degree $n$ and $d (x)$ of degree $m$ with leading coefficient $b_{m}$ .

If $de g p < de g d$ , take $q (x) = 0$ and $r (x) = p (x)$ . Done.
Otherwise ( $n \geq m$ ), form the term $t (x) = \frac{a _{n}}{b _{m}} x^{n - m}$ .
The product $t (x) d (x)$ has leading term $\frac{a _{n}}{b _{m}} x^{n - m} \cdot b_{m} x^{m} = a_{n} x^{n}$ , matching the leading term of $p$ .
The difference $p_{1} (x) = p (x) - t (x) d (x)$ therefore has the $x^{n}$ term cancelled, so $de g p_{1} < n$ .
Repeat steps 2–4 on $p_{1}$ , accumulating the $t$ -terms into the quotient $q (x)$ . Since each step strictly lowers the degree, after finitely many steps the running remainder has degree $< m$ . Call it $r (x)$ .
Summing the accumulated relations gives $p (x) = d (x) q (x) + r (x)$ with $de g r < de g d$ . $■$

Uniqueness: Suppose

p = d q_{1} + r_{1} = d q_{2} + r_{2}

. Subtracting,

d (q_{1} - q_{2}) = r_{2} - r_{1}

. The right side has degree

< m

; the left side, if

q_{1} \neq = q_{2}

, has degree

\geq m

— a contradiction unless both sides are zero. Hence

q_{1} = q_{2}

and

r_{1} = r_{2}

Évariste Galois | 1811–1832

Though polynomial arithmetic is ancient, the deepest understanding of which polynomials can be solved by radicals came from Galois, who died in a duel at 20. He showed that to each polynomial equation one can attach a group of symmetries of its roots, and that solvability by formula corresponds to a special ("solvable") structure of that group. This explained why the quintic ( $de g 5$ ) has no general radical formula while degrees 1–4 do, founding modern algebra in the process — built directly atop the arithmetic of polynomials described here.

Building Intuition

Think of a polynomial as a base- $x$ number whose "digits" are the coefficients, except the digits can be any size and even negative. Then:

Adding polynomials = adding digit-by-digit with no carrying (because $x$ is a symbol, not the fixed value 10).
Multiplying polynomials = exactly the grade-school multiplication grid, again without carries — each diagonal of the grid becomes one coefficient $c_{m}$ .
Dividing polynomials = long division you already know, lined up by powers of $x$ .

The absence of carrying is precisely what makes polynomial arithmetic cleaner than integer arithmetic, and why algorithms (and the roots that come from setting $p (x) = 0$ ) behave so predictably. The graph of a polynomial is a smooth, unbroken curve whose degree caps how many times it can cross the $x$ -axis and how many "wiggles" (turning points, at most $n - 1$ ) it can have.

Graphs of polynomials by degree: each crosses the x-axis at most "degree" times.

Worked Examples

Multiplying and Collecting Like Terms

Given:

p (x) = 2 x^{2} - 3 x + 1

and

q (x) = x^{2} + 4 x - 5

. Find

p (x) \cdot q (x)

. Steps — apply [3], multiplying each term of

p

by each term of

q

| | $x^{2}$ | $+ 4 x$ | $- 5$ | |---|---|---|---| | $2 x^{2}$ | $2 x^{4}$ | $8 x^{3}$ | $- 10 x^{2}$ | | $- 3 x$ | $- 3 x^{3}$ | $- 12 x^{2}$ | $15 x$ | | $+ 1$ | $x^{2}$ | $4 x$ | $- 5$ |

Collect by power (the convolution $c_{m} = \sum_{i + j = m} a_{i} b_{j}$ ):

$x^{4}$ : $2$
$x^{3}$ : $8 - 3 = 5$
$x^{2}$ : $- 10 - 12 + 1 = - 21$
$x^{1}$ : $15 + 4 = 19$
$x^{0}$ : $- 5$

Degree check via [4]: $de g p + de g q = 2 + 2 = 4$ . ✓

p (x) \cdot q (x) = 2 x^{4} + 5 x^{3} - 21 x^{2} + 19 x - 5

Long Division with Remainder, then Confirming a Root

Given: Divide

p (x) = 2 x^{3} - 5 x^{2} + 3 x - 4

d (x) = x - 2

, then state

p (2)

. Steps (polynomial long division, lining up powers):

$2 x^{3} \div x = 2 x^{2}$ . Subtract $2 x^{2} (x - 2) = 2 x^{3} - 4 x^{2}$ : remainder $- x^{2} + 3 x - 4$ .
$- x^{2} \div x = - x$ . Subtract $- x (x - 2) = - x^{2} + 2 x$ : remainder $x - 4$ .
$x \div x = 1$ . Subtract $1 \cdot (x - 2) = x - 2$ : remainder $- 2$ .

So $q (x) = 2 x^{2} - x + 1$ and $r = - 2$ :

2 x^{3} - 5 x^{2} + 3 x - 4 = (x - 2) (2 x^{2} - x + 1) - 2

By the Remainder Theorem [6], the remainder equals $p (2)$ . Check directly: $p (2) = 2 (8) - 5 (4) + 3 (2) - 4 = 16 - 20 + 6 - 4 = - 2$ . ✓

Because $r = - 2 \neq = 0$ , $(x - 2)$ is not a factor; $x = 2$ is not a root.

q (x) = 2 x^{2} - x + 1, r = - 2 = p (2)

Common Misconceptions

Pitfalls that quietly produce wrong answers:

$x$ , $x^{- 1}$ , $2^{x}$ , and $\frac{1}{x + 1}$ are not polynomials — exponents must be non-negative integers, and the variable may not sit in a denominator or under a root.
$(p + q)^{2} \neq = p^{2} + q^{2}$ . The cross term matters: $(a + b)^{2} = a^{2} + 2 ab + b^{2}$ .
"Degree of a sum equals the max degree" can fail when leading terms cancel — that is why [4] uses $\leq$ for sums.
When dividing, you must include placeholder zeros for missing powers (e.g. write $x^{3} + 0 x^{2} + 0 x - 1$ ), or the columns misalign.
A polynomial of degree $n$ has at most $n$ real roots, but possibly fewer real roots (some may be complex or repeated).

Claim: "Adding two degree-3 polynomials gives a degree-3 polynomial." Take $p (x) = x^{3} + x$ and $q (x) = - x^{3} + 2$ . Then

p (x) + q (x) = x + 2,

which has degree 1, not 3 — the leading terms cancelled. This is exactly why the degree rule for addition in [4] is an inequality, never an equality.

Connections

Polynomial arithmetic is a hub connecting many areas:

The set of polynomials forms a ring (closed under $+, -, \times$ with the [5] division algorithm), structurally parallel to the integers — see Ring Theory and Euclidean Domains.
Setting $p (x) = 0$ leads to Roots of Polynomials and the Fundamental Theorem of Algebra.
Breaking polynomials into irreducible pieces is Factoring Polynomials, the analog of prime factorization.
The special case $de g 2$ is governed by The Quadratic Formula and Completing the Square.
Derivatives and integrals of polynomials are the easiest cases in Differential Calculus and connect to Taylor Series, where general functions are approximated by polynomials.
Multiplication-as-convolution [3] links directly to Discrete Convolution and the Fast Fourier Transform used to multiply huge polynomials quickly.

Real-World Applications

Computer graphics & fonts — Bézier curves and splines (piecewise cubic polynomials) define the smooth outlines of letters in TrueType/PostScript fonts and the motion paths in animation software.
Error-correcting codes — Reed–Solomon codes (used in QR codes, CDs/DVDs, and deep-space probes like Voyager) treat data as polynomial coefficients; remainders under polynomial division detect and repair errors.
Cryptography — Shamir's Secret Sharing splits a secret into shares that are points on a polynomial, recoverable only when enough shares are combined (via the Factor/interpolation idea).
Numerical interpolation — Engineers fit polynomials through measured data points (e.g. sensor calibration curves, temperature-vs-resistance tables).
Physics of motion — Projectile height $h (t) = - \frac{1}{2} g t^{2} + v_{0} t + h_{0}$ is a quadratic; its roots give launch and landing times.
Economics & optimization — Cost, revenue, and profit functions are frequently modeled as polynomials whose turning points (max profit) are found by calculus.
Signal processing — Digital filters are specified by the coefficients of polynomials in the variable $z$ , and their roots ("poles and zeros") determine the filter's behavior.

Historical Development

c. 1650 BCE — The Egyptian Rhind Papyrus and Babylonian clay tablets solve quadratic-type problems geometrically, without symbolic notation.
c. 250 CE — Diophantus of Alexandria's Arithmetica introduces syncopated (abbreviated) notation for unknowns and powers — a step toward polynomial symbolism.
c. 820 CE — Al-Khwārizmī's al-Jabr systematizes solving linear and quadratic equations, giving algebra its name and its method of "balancing."
1545 — Gerolamo Cardano publishes Ars Magna with general solutions to the cubic (from Tartaglia/del Ferro) and quartic (Ferrari), pushing arithmetic of higher-degree polynomials.
1591 — François Viète introduces letters for both knowns and unknowns, the birth of true symbolic algebra and coefficient manipulation.
1637 — René Descartes' La Géométrie fixes much of our modern notation ( $x^{2}$ , $x^{3}$ ) and links polynomials to curves via coordinates.
1799 — Carl Friedrich Gauss gives the first rigorous proof of the Fundamental Theorem of Algebra: every degree- $n$ polynomial has $n$ complex roots.
1824 / 1832 — Niels Abel proves the general quintic is unsolvable by radicals; Évariste Galois explains why via group theory.
1965 — Cooley & Tukey publish the Fast Fourier Transform, dramatically speeding up the multiplication of large polynomials (and integers) in computing.

Subtleties & Limits

Several deep questions about polynomials remain hard or open:

Factoring difficulty: While factoring polynomials over the rationals is algorithmically solvable, integer factorization (a special multivariate-style problem) underlies RSA cryptography, and its hardness is unproven.
Sparse polynomials: Detecting whether a polynomial given in compact "sparse" form is identically zero, or counting its real roots efficiently, raises subtle complexity questions still under study.
Hilbert's Tenth Problem (resolved 1970): There is no algorithm to decide whether an arbitrary multivariable polynomial equation has integer solutions — a striking limit on what polynomial arithmetic can ultimately settle.
Open: Sendov's conjecture (proved for high degree only recently) and questions on the distribution of roots of random polynomials remain active research.