Polynomials over $ \mathbb{Z}/p\mathbb{Z}{}$

The polynomials $ x^2-1$ has four roots in $ \mathbb{Z}/8\mathbb{Z}{}$ , namely $ 1$ , $ 3$ , $ 5$ , and $ 7$ . In contrast, the following proposition shows that a polynomial of degree $ d$ over a field, such as $ \mathbb{Z}/p\mathbb{Z}{}$ , can have at most $ d$ roots.

Proposition 2.5 (Root Bound)   Let $ f\in k[x]$ be a nonzero polynomial over a field $ k$ . Then there are at most $ \deg(f)$ elements $ \alpha\in k$ such that $ f(\alpha)=0$ .

Proof. We prove the proposition by induction on $ \deg(f)$ . The cases in which $ \deg(f)\leq 1$ are clear. Write $ f = a_n x^n + \cdots a_1 x + a_0$ . If $ f(\alpha)=0$ then

$\displaystyle f(x)$ $\displaystyle = f(x) - f(\alpha)$    
  $\displaystyle = a_n(x^n-\alpha^n) + \cdots + a_1(x-\alpha) + a_0(1-1)$    
  $\displaystyle = (x-\alpha)(a_n(x^{n-1}+\cdots + \alpha^{n-1}) + \cdots + a_2(x+\alpha) + a_1)$    
  $\displaystyle = (x-\alpha)g(x),$    

for some polynomial $ g(x)\in k[x]$ . Next suppose that $ f(\beta)=0$ with $ \beta\neq \alpha$ . Then $ (\beta-\alpha) g(\beta) = 0$ , so, since $ \beta-\alpha\neq 0$ and $ k$ is a field, we have $ g(\beta)=0$ . By our inductive hypothesis, $ g$ has at most $ n-1$ roots, so there are at most $ n-1$ possibilities for $ \beta$ . It follows that $ f$ has at most $ n$ roots. $ \qedsymbol$

SAGE Example 2.5   We use SAGE to find the roots of a polynomials over $ \mathbb{Z}/13\mathbb{Z}{}$ .
sage: R.<x> = PolynomialRing(Integers(13))
sage: f = x^15 + 1
sage: f.roots()
[(12, 1), (10, 1), (4, 1)]
sage: f(12)
0
The output of the roots command above lists each root along with its multiplicity (which is 1 in each case above).

Proposition 2.5   Let $ p$ be a prime number and let $ d$ be a divisor of $ p-1$ . Then $ f = x^d-1\in(\mathbb{Z}/p\mathbb{Z}{})[x]$ has exactly $ d$ roots in $ \mathbb{Z}/p\mathbb{Z}{}$ .

Proof. Let $ e=(p-1)/d$ . We have

$\displaystyle x^{p-1} - 1$ $\displaystyle = (x^d)^e - 1$    
  $\displaystyle = (x^d - 1)((x^d)^{e-1} + (x^d)^{e-2} + \cdots + 1)$    
  $\displaystyle = (x^d - 1)g(x),$    

where $ g\in (\mathbb{Z}/p\mathbb{Z}{})[x]$ and $ \deg(g) = de-d = p-1-d$ . Theorem 2.1.19 implies that $ x^{p-1}-1$ has exactly $ p-1$ roots in $ \mathbb{Z}/p\mathbb{Z}{}$ , since every nonzero element of $ \mathbb{Z}/p\mathbb{Z}{}$ is a root! By Proposition 2.5.3$ g$ has at most $ p-1-d$ roots and $ x^d-1$ has at most $ d$ roots. Since a root of $ (x^d-1)g(x)$ is a root of either $ x^d-1$ or $ g(x)$ and $ x^{p-1}-1$ has $ p-1$ roots, $ g$ must have exactly $ p-1-d$ roots and $ x^d-1$ must have exactly $ d$ roots, as claimed. $ \qedsymbol$

SAGE Example 2.5   We use SAGE to illustrate the proposition.
sage: R.<x> = PolynomialRing(Integers(13))
sage: f = x^6 + 1
sage: f.roots()
[(11, 1), (8, 1), (7, 1), (6, 1), (5, 1), (2, 1)]

We pause to reemphasize that the analogue of Proposition 2.5.5 is false when $ p$ is replaced by a composite integer $ n$ , since a root mod $ n$ of a product of two polynomials need not be a root of either factor. For example, $ f = x^2-1=(x-1)(x+1)\in \mathbb{Z}/15\mathbb{Z}{}[x]$ has the four roots $ 1$ , $ 4$ , $ 11$ , and $ 14$ .

William 2007-06-01