Matching Roots, Turning Points, and Multiplicity

Why Roots and Turning Points Must Match

A polynomial’s graph cannot combine roots, crossings, and turning points arbitrarily. Its degree limits how many times the graph can meet the horizontal axis and how many distinct places its derivative can be zero. Root multiplicities impose additional restrictions: they determine whether the graph crosses the axis or touches it and turns.

These restrictions are useful when sketching a polynomial and when checking whether a proposed sketch could belong to a polynomial of a specified degree. The key is to count carefully: a repeated root uses up degree, and a turning point requires a zero of the derivative.

Roots, Multiplicity, and Crossing Behavior

Near \(r\), the factor \(q(x)\) keeps the same sign as \(q(r)\), since it is continuous and nonzero there. The sign of \(p(x)\) on either side of \(r\) is therefore controlled by \((x-r)^m\). If \(m\) is even, that factor has the same sign on both sides: the graph touches the axis but does not cross it. If \(m\) is odd, its sign changes: the graph crosses the axis.

Multiplicity also says something about the slope at the root. When \(m=1\), the root is simple and \(p'(r)\ne 0\). When \(m\ge 2\), \(p'(r)=0\), so the root is a critical point. But not every critical point is a turning point: at an odd-multiplicity root such as a triple root, the graph can cross the axis while continuing in the same direction.

Degree Limits the Number of Turning Points

A nonconstant polynomial of degree \(n\) has a derivative of degree \(n-1\). A nonzero polynomial of degree \(n-1\) has at most \(n-1\) distinct real roots. Every local maximum or minimum of a differentiable function at an interior point of its domain has derivative zero. Consequently, a degree-\(n\) polynomial can have at most \(n-1\) distinct turning points.

This is an upper bound, not a promise that all \(n-1\) points occur. A derivative zero can fail to be a turning point if the derivative has the same sign on both sides. Also, if a domain is restricted, a boundary extremum need not have derivative zero; here, “turning point” means a local maximum or minimum at an interior point of the polynomial’s domain, which is all of \(\mathbb R\).

How Roots Force Critical Points Between Them

Distinct real roots impose another constraint. If \(a<b\) are consecutive distinct real roots of a polynomial, then the polynomial has no root in \((a,b)\). It has a constant, nonzero sign throughout that interval. Since its value is zero at both endpoints, it must attain a positive maximum inside the interval if it is positive there, or a negative minimum inside if it is negative there. That interior extremum is a turning point. Equivalently, Rolle’s theorem guarantees at least one derivative zero in \((a,b)\).

Even-multiplicity roots give turning points at the roots themselves, while consecutive distinct roots force a turn strictly between them. These locations do not overlap: the roots are endpoints of the intervals, and the forced turns are inside them. Thus, if a polynomial has \(k\) distinct real roots and \(e\) of those roots have even multiplicity, it has at least \(k-1+e\) distinct turning points. In particular, this count cannot exceed \(n-1\).

Multiplicity also gives a more detailed derivative count. If the distinct real roots are \(r_1,\ldots,r_k\), with multiplicities \(m_1,\ldots,m_k\), then each root \(r_i\) contributes a derivative root of multiplicity \(m_i-1\) when \(m_i\ge2\). Rolle’s theorem supplies at least one additional derivative root in each of the \(k-1\) open intervals between consecutive roots. Together these account for at least \(\sum_{i=1}^{k}(m_i-1)+(k-1)=\sum_{i=1}^{k}m_i-1\) roots of the derivative, counted with multiplicity. Since the derivative has degree \(n-1\), this total cannot exceed \(n-1\).

Why Multiplicity Determines the Local Shape

Write a polynomial near a root \(r\) of multiplicity \(m\) as \(p(x)=(x-r)^m q(x)\), where \(q(r)\ne0\). Differentiating gives \(p'(x)=m(x-r)^{m-1}q(x)+(x-r)^m q'(x)\). Factoring out the common power yields:

The bracketed factor has value \(mq(r)\ne0\) at \(r\). Thus \(p'\) has a root of multiplicity exactly \(m-1\) at \(r\). If \(m\) is even, then \(m-1\) is odd, so \(p'\) changes sign there and the polynomial turns. If \(m\) is odd and at least three, then \(m-1\) is even, so \(p'\) does not change sign there: the graph has a horizontal tangent but does not turn.

Testing Whether a Proposed Sketch Is Possible

A proposed feature list should be checked against both the degree bound and the root pattern. For example, four distinct roots already force at least three turns between consecutive roots. If one of those roots has even multiplicity, it contributes an additional turn at the root. The maximum-turn bound may then leave no room for any extra turns elsewhere.

A Common Mistake: Treating Every Critical Point as a Turn

A frequent error is to see \(p'(r)=0\) and conclude that the graph must have a local maximum or minimum at \(r\). A zero derivative is necessary at an interior turn, but it is not sufficient. For a concrete counterexample, let \(p(x)=(x-1)^3\). Then \(p'(x)=3(x-1)^2\), so \(p'(1)=0\). But \(p'(x)>0\) on both sides of \(1\), and \(p(x)\) changes from negative to positive there. The graph crosses with a horizontal tangent and does not turn.

The correction is to check how the derivative behaves on either side of the critical point. A change from positive to negative indicates a local maximum; a change from negative to positive indicates a local minimum. If there is no sign change, the critical point is not a turning point. For polynomial roots, multiplicity predicts this behavior: even multiplicity gives a turn, while odd multiplicity gives a crossing.

Using the Constraints in Practice

When checking a polynomial sketch, keep three counts separate: distinct roots, total root multiplicity, and distinct turning points. A degree-\(n\) polynomial has at most \(n\) roots counted with multiplicity and at most \(n-1\) distinct turning points. Its consecutive distinct real roots force at least one turn in each gap, and each even-multiplicity root adds a turn at that root. Odd roots of multiplicity at least three are stationary crossings, not turns.

These are necessary constraints, not a complete construction method. Passing the counts does not by itself prove that a polynomial with every proposed feature exists. But failing any one of them proves that the proposed combination is impossible. That makes degree and multiplicity a useful first consistency check before attempting a detailed sketch.