255 Sol 2

Problem I.
Let ρ denote the reduction map R[x] → R/I[x], and let π denote the reduction map R → R/I. Let
f, g ∈ R[x]. Write f = an xn + · · · + a0 and g = bn xn + · · · + b0 where ai , bi ∈ R (for simplicity we
write out the same number of coefficients for both f and g, but they can still have different degrees
if an = 0). Then ρ(f + g) = ρ((an + bn )xn + · · · + (a0 + b0 )) = π(an + bn )xn + · · · + π(a0 + b0 ) =
π(an )xn + π(bn )xn + · · · + π(a0 ) + π(b0 ) = π(an )xn + · · · + π(a0 ) + π(bn )xn + π(b0 ) = ρ(f ) + ρ(g),
where we use the fact that ρ is a ring homomorphism. Similarly, we find that ρ(f g) = ρ(f )ρ(g),
because we can just apply π to the coefficients of f g to get ρ(f g) and then use the fact that π is a
homomorphism and factor into ρ(f )ρ(g).
Problem II.
Suppose rs is a rational root of p(x) ∈ Z[x] where gcd(r, s) = 1. Write p(x) = an xn + an−1 xn−1 +
n n−1 n
· · · + a0 where an 6= 0. Then we have that 0 = an rs + an−1 rs + · · · + a0 = an rsn +
n−1
an−1 rsn−1 + · · · + a0 . Multiplying through by sn , we get an rn + an−1 rn−1 s + · · · + a0 sn = 0. Then
an rn + an−1 rn−1 s + · · · + a1 rsn−1 = −a0 sn . Because r divides the left side, we have r|a0 sn . Because
gcd(r, s) = 1, this means that r|a0 . Similarly, by subtracting off an rn instead, we find that s|an .
This is what we wanted.
Problem III.
Let f be a monic polynomial with integer coefficients. Clearly if f is irreducible in Q[x], it is also
irreducible in Z[x], because a factorization in Z[x] would give a factorization in Q[x]. Note this we
use monicity here: if g is monic and irreducible in Z[x], then f = 2g is not reducible in Z[x], but
it is in Q[x] because 2 is a unit in Q[x]. However, ensuring that f is monic takes care of this issue.
Now suppose that f is reducible in Q[x]. Then f = ab for some a, b ∈ Q[x] of degree at least 1.
Write a = pqnn xn + · · · + pq00 where gcd(pi , qi ) = 1 for each i. Let l1 = lcm(q0 , . . . , qn ). Then l1 a is a
polynomial with integer coefficients whose gcd is 1. We can define a number l2 for b in the same
way and find that l2 b is an also a polynomial with integer coefficients whose gcd is 1. Then, by
Gauss’ Lemma, l1 al2 b is a polynomial with integer coefficients whose gcd is 1. But l1 al2 b = l1 l2 f .
If l1 l2 6= 1, this is a contradiction because l1 l2 divides all the coefficients of l1 l2 f . Thus we must
have l1 l2 = 1 which means l1 = l2 = 1. Therefore, all of the denominators in a and b are 1, so a
and b actually have integer coefficients. Thus f = ab is a factorization in Z[x], so f is reducible in
Z[x]. Thus f is irreducible in Z[x] iff it is irreducible in Q[x].
Problem IV.
(1) Suppose f is reducible in Q[x]. By the previous problem, this means that f is reducible in Z[x].
(Even though we don’t have the condition that f is monic, we’re still ok—the only place in the
proof of III that we used monicity was to make sure that we couldn’t just factor out an integer
coefficient. But if f is irreducible in Q[x], then we don’t have this problem and the proof still
works.) Then f = ab with a, b ∈ Z[x] both of degree at least 1. Then π(f ) = π(a)π(b). Let
a0 , b0 ∈ Z be the leading coefficients of a and b respectively. Then a0 b0 is the leading coefficient
of f . Because p doesn’t divide a0 b0 , p doesn’t divide a0 or b0 . This means that the leading terms
of a don’t go away when we apply π, so π(a)π(b) is a factorization of π(f ). Thus, if π(f ) is
irreducible, then f is irreducible in Z[x].
(2) x2 + p irreducible in Z[x] by Eisenstein’s criterion. But π(x2 + p) = x2 which is reducible.
1
Problem V.
We know that Φp (x)(x − 1) = xp − 1. Thus Φp (x + 1)((x + 1) − 1) = Φp (x + 1)x = (x + 1)p − 1 =

xp + pxp−1 + p2 xp−2 + · · · + px + 1 − 1. Thus Φp (x + 1) = xp−1 + pxp−2 + p2 xp−3 + · · · + p. By

Eisenstein’s criterion, Φp (x + 1) is irreducible. This means that Φp (x) is irreducible (a factorization

could be turned into a factorization of Φ(x + 1)).
Problem VI.
We follow the Euclidean algorithm:

(x4 − x3 − x2 + 1) − x(x3 − 1) = −x3 − x2 + x + 1
(−x3 − x2 + x + 1) + (x3 − 1) = −x2 + x
(x3 − 1) + x(−x2 + x) = x2 − 1
(−x2 + x) + (x2 − 1) = x − 1
(x2 − 1) − (x + 1)(x − 1) = 0
So the gcd is x − 1. We can write this as a linear combination of the original two polynomials
f = x4 − x3 − x2 + 1 and g = x3 − 1 by looking at the steps we followed and working backwards:
x − 1 =(−x2 + x) + (x2 − 1)
=((−x3 − x2 + x + 1) + g) + (g + x(−x2 + x))
=((f − xg) + g) + (g + x((−x3 − x2 + x + 1) + g))
=f − xg + g + g + x(f − xg + g)
=(x + 1)f + (−x2 + 2)g
The other two parts can be done in the same way. In (2), be careful to remember that coefficients
are all mod 5.
Problem VII.
(1) Suppose p = ab, with a, b ∈ R. Then ab ∈ (p), and because (p) is a prime ideal, a ∈ (p) or
b ∈ (p). Suppose WLOG that a ∈ P . Then a = px for some x ∈ R. Then pxb = p. Because R
is an integral domain, we have the cancellation property, so xb = 1. Thus b has a multiplicative
inverse, so b ∈ R× . Therefore p is irreducible.
(2) Because Euclidean domains are integral domains, if (p) is prime, then p is irreducible by part
(1). Now suppose p 6= 0 is irreducible. Let I be an ideal with (p) ⊆ I ⊆ R. Because every ideal
is principal, there’s some q ∈ R with (q) = I. Then p ∈ (q) so p = qx for some x ∈ R. But
p is irreducible, so we must have x ∈ R× . But then q = px−1 , so q ∈ (p), so (q) ⊆ (p). Thus
(p) = (q). This means that (p) is a maximal ideal, so it is therefore also a prime ideal.
Problem VIII.
Suppose p is irreducible in F [x]. Now pick a ∈ F [x] with a ∈ / (p). Then gcd(a, p) = 1, so
by the Euclidean algorithm we can find f, g ∈ F [x] such that af + px = 1. Then in F [x]/(p),
(a + (p))(f + (p)) = 1 + (p), so a + (p) is invertible. Thus every nonzero element of F [x]/(p) is
invertible, so F [x]/(p) is a field.
Suppose F [x]/(p) is a field. Then for each a ∈ F [x], a ∈ / (p), a + (p) is invertible in F [x]/(p),
i.e. there’s some b ∈ F [x] with (a + (p))(b + (p)) = 1 + (p) in F [x]/(p). This means that ab + pg = 1
2
for some g ∈ F [x]. Let a, b ∈ F [x] such that ab = p. If a|p, then b must be a unit. Suppose
that a doesn’t divide p. Then by the above, there’s some c and g with ac + pg = 1. Then
abc + pgb = pc + pgb = b. Thus p|b, which means that a is a unit. Thus p is irreducible.
Problem IX.
It’s straightforward to check that R is a subring. Now, any elements of R of degree 2 or 3 are
irreducible in R: if such an element can be factorized, one of the factors must be linear, but there
are no linear elements of R. In particular, x2 and x3 are irreducible in R. Thus (x2 )3 = (x3 )2 give
two distinct factorizations of x6 into irreducibles.
Problem X.
Let pn = (x−1)(x−2) . . . (x−n)−1. Because pn is monic, by problem III it’s sufficient to check that
pn is irreducible over Z. Suppose pn is reducible, so there is f, g ∈ Z[x] with 1 ≤ deg f, deg g < n.
Note that pn (k) = −1 for k = 1, 2, . . . , n. Then, because f (k), g(k) ∈ Z when k ∈ Z, we must have
that for each k = 1, 2, . . . , n, f (k) = ±1 and g(k) = −f (k). Thus f (k) + g(k) = 0, so f + g has
roots at 1, 2, . . . , n. But f + g is a polynomial of degree less than n, and thus has fewer than n
roots. This is a contradiction. So pn is irreducible.

255 Sol 2

Uploaded by

Copyright:

Available Formats

255 Sol 2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

255 Sol 2

Uploaded by

Copyright:

Available Formats

Problem I.

(2) x2 + p irreducible in Z[x] by Eisenstein’s criterion. But π(x2 + p) = x2 which is reducible.

We know that Φp (x)(x − 1) = xp − 1. Thus Φp (x + 1)((x + 1) − 1) = Φp (x + 1)x = (x + 1)p − 1 =

Eisenstein’s criterion, Φp (x + 1) is irreducible. This means that Φp (x) is irreducible (a factorization

We follow the Euclidean algorithm:

You might also like