Polynomials: The Grammar of Curved Expressions

The graphing calculator settled an argument in four seconds that words had failed to settle in four minutes. My daughter insisted x cubed and x squared were basically the same shape, bigger versus smaller, and I insisted they were different animals. We graphed both: the parabola smiled symmetrically upward while the cubic swooped down from the left, flattened through the middle, and climbed away to the right. Same family, different temperaments, and the difference had a name printed nowhere on the screen: degree. Three versus two, and everything about the ending changes.

Polynomials are sums of whole-number powers of x with number coefficients, and the highest power, the degree, runs the show. It decides how many bends the graph can have, where both ends point, and how many solutions hide inside. Five sections below read that grammar the way you read a sentence diagram: degree first, then ends, then roots, then the factoring patterns that crack them open, and finally division for the leftovers.

Degree runs the show

Every polynomial introduces itself by degree: the highest power present. A constant has degree zero and graphs flat. x plus anything is degree one, a line, the only honest shape in the family. Our expressions guide teaches the vocabulary, and linear equations own degree one completely. From degree two upward, curves begin, and each new degree buys exactly one more bend the graph is allowed to make.

Quadratics smile or frown and stop there: one bend, one valley or peak, no second thoughts. Our quadratics guide runs the full shape. Cubics bend twice at most, drawing the sideways S that fooled my daughter into thinking it was a parabola with ambition. Quartics manage three bends, quintics four, and the rule holds all the way up: degree n allows at most n minus 1 turns. Count the bends on any sketch and you have an upper bound on the degree for free.

The leading coefficient, the number in front of the highest power, votes alongside degree. Positive leading numbers push right ends upward; negative ones drag them down. Combined with degree, two numbers predict the whole silhouette: even degree with a positive lead rises on both sides like a smile with extra wiggles, odd degree with a positive lead falls left and climbs right like a road over a hill. Sketch both archetypes from memory before meeting any specific polynomial, and the particulars will hang on hooks you already own.

Ends tell the story first

End behavior is the cheapest prediction in algebra: ignore everything except the highest-degree term and ask where the graph goes at both horizons. For 3x to the fourth minus a hundred x, the x-to-the-fourth term wins everywhere far out, and since the degree is even with a positive coefficient, both ends climb. The middle may wiggle, cross, and loop, but the ends are decided before the first calculation finishes, which is why sketching starts at the edges and works inward.

Odd degrees guarantee a crossing, and that guarantee is worth money. A cubic falling on the left and climbing on the right must pass through zero somewhere in between, continuous curve, no jumps allowed, so every odd-degree polynomial has at least one real root. Job interviewers for engineering roles have asked exactly this as a reasoning probe: not the root, just whether one must exist. The answer is yes, always, by the shape of the ends alone.

Practice the read on four specimens: x squared (up-up smile), negative x cubed (up-left, down-right frown-slide), x to the fourth (up-up with wiggles allowed), negative x to the fifth (down-right slide reversed). Name degree and sign, predict both ends, then graph to check. Four correct predictions in a row and the skill is installed, because end behavior is pattern matching with exactly two inputs.

Roots are where the curve meets the ground

Every factor of the form x minus a number hands you a root for free: set it to zero and x equals a, which is exactly where the graph crosses the horizontal axis. A degree-n polynomial splits into n such factors, counting repeats, which means it hides at most n real crossings, and exactly n roots if you count complex ones too (Wikipedia, Fundamental theorem of algebra). The degree is a budget: spend it on crossings, one per factor, and the account cannot overdraw.

Repeated factors behave differently at the axis, and the difference is visible. A single factor crosses cleanly through; a squared factor, like x minus 2 squared, touches the axis and bounces back, tangent rather than crossing. Teach the eye first: cross means odd multiplicity, kiss means even, and the graph testifies about the algebra without solving anything. Bounces are not near-misses. They are even multiplicities announcing themselves.

Work one fully. Factor x squared minus 5x plus 6 into x minus 2 times x minus 3, read off roots two and three, and sketch: smile shape, crossings at 2 and 3, valley between them. Three facts, one factorization, zero computation beyond splitting 5 into 2 and 3. Factoring converts equation-solving into reading, which is why the skill pays rent across every later course.

Three patterns that crack most cases

Always pull the common factor first, before any cleverness. Six x cubed plus nine x squared surrenders 3x squared immediately, leaving 2x plus 3, and half of all factoring failures are skipped extractions rather than missed patterns. GCF first is not a suggestion; it is the toll booth every factorization passes through, and skipping it strands the work two steps later.

Difference of squares splits anything shaped a squared minus b squared into a minus b times a plus b, no exceptions and no thinking required. Nine x squared minus 16 is 3x minus 4 times 3x plus 4, spotted in one glance once the pattern lives in muscle memory. Sum and difference of cubes follow the same instinct with longer templates: a cubed plus b cubed is a plus b times the quadratic with flipped sign and no middle cooperation. Memorize the three shapes the way faces are memorized, and factoring becomes recognition instead of invention.

Our exponents guide stores the power rules these patterns spend. A student fluent in exponents sees x to the fourth minus 81 as squares of squares and reaches for difference-of-squares twice, landing on four linear factors without a single guess. Fluency upstream is speed downstream, and factoring speed is almost entirely recognition speed.

Dividing what won't factor

Some polynomials refuse every pattern, and division is the honest fallback. Polynomial long division mirrors the arithmetic kind digit for digit: divide the leading terms, multiply back, subtract, bring down the next term, repeat until the remainder's degree drops below the divisor's. Our long division guide walks the machine. The only new rule is sorting by descending powers first, because the algorithm reads left to right and unsorted input scrambles it.

Synthetic division is the shortcut when the divisor is linear, x minus a number: bring down coefficients, multiply diagonally, add down the column, and the bottom row is the quotient with the remainder at the end. One row of numbers replaces a page of algebra, and the remainder doubles as a function evaluation, the value of the polynomial at a, which turns checking into arithmetic. Learn it after long division, never before, because shortcuts teach nothing to hands that never did the long way.

The remainder is the diagnostic. Zero remainder means the divisor was a factor and the polynomial splits further; nonzero remainder names exactly what is left over, which the next section of the course will use. Division doesn't just compute. It interrogates, and the remainder is the answer it extracts under pressure.

The one habit: Friday degree

Every Friday, take one expression from the week and file it: degree, leading coefficient, predicted ends, actual sketch. Four lines in the notebook, two minutes, and the filing habit turns every new polynomial into a familiar animal. My daughter's file holds forty entries now, and she reads new ones the way birders read silhouettes: degree first, ends second, roots on demand. Fifty-two Fridays a year is a field guide written by its own reader, and no curve surprises its author twice. Retired entries graduate to a master list, and the master list becomes the fastest review sheet in the house.