Selected exercises from Abstract Algebra by Dummit

and Foote (3rd edition).

Bryan Félix
Abril 12, 2017

Section 10.1

Exercise 8. An element m of the R-module M is called a torsion element if rm = 0 for some

nonzero element r ∈ R. The set of torsion elements is denoted

Tor(M ) = {m ∈ M | rm = 0 for some non zero r ∈ R}

(a) Prove that if R is an integral domain then Tor(M ) is a submodule of M .

Proof. We use the submodule criterion. Observe that since R is an integral domain 1 6= 0
and 1 · 0 = 0. Hence Tor(M ) is non empty. Now let x, y be in Tor(M ) so that r1 x = 0
and r2 y = 0; and let r be an arbitrary element of R. Then (using the fact that R is
commutative), observe that

r1 r2 (x + ry) = r2 (r1 x) + r1 r(r2 y) = 0 + 0 = 0.

Hence x + ry is in the torsion as we wanted.

(b) Give an example of a ring R and an R-module M such that Tor(M ) is not a submodule.

Solution. Taking a hint from part (a) we want R to not be an integral domain. Let
R = Z/10Z and consider the R-module generated by itself. Note that 2 and 5 are torsion
elements as 2 · 5 ≡ 0 mod 10, but 7 = 2 + 5 is not, since r · 7 = 0 has only the trivial
solution. Hence Tor(M ) is not closed under addition, and therefore is not a submodule.

(c) If R has zero divisors show that every nonzero R-module has non zero torsion elements.

Proof. Let M be a nonzero R-module. Let a and b be the zero divisor elements in R and
take m 6= 0 ∈ M (m exists since M is non zero). Then, if b · m = 0 we are done, and m is
in the torsion. Otherwise a(b · m) = (ab) · m = (0) · m = 0, and (b · m) is in the torsion as
desired.

Exercise 12. In the notation of the preceding exercises prove the following facts about annihi-
lators.

(a) Let N be the submodule of M and let I be its annihilator in R. Prove that the annihilator
of I contains N . Give an example where the annihilator of I in M does not equal N .

1
 Proof. Let n ∈ N and consider the product n · i where i is an arbitrary element in I. Since
i is an element of the annihilator of N we see that n · i = 0 as needed.
Let Z/2Z×Z/2Z be a Z-module and let Z/2Z×0 be a submodule. Let I = Ann(Z/2Z×0) =
2Z and note that Ann(I) = Z/2Z × Z/2Z 6= Z/2Z × 0.

(b) Let I be a right ideal of R and let N be its annihilator in M . Prove that the annihilator of
N in R contains I. Give an example where the annihilator of N in R does not equal I.

Proof. Let i be an element of I, and consider the product n · i where n is an arbitrary

element of N . Since n is in the annihilator of I in M . Then n · i = 0 as required.
Let Z/2Z × Z/6Z be a Z-module, and let i = 4Z be a right ideal. Note that Ann(4Z) (in
Z/2Z × Z/6Z) is the set

{(a, b) ∈ Z/2Z × Z/6Z | 4na = 0 mod 2, 4nb = 0 mod 6 for all n}

Thus,
Ann(4Z) = {(0, 0), (0, 3), (1, 0), (1, 3)}.
Then we inspect the annihilator in Z of {(0, 0), (0, 3), (1, 0), (1, 3)}. Note that in order to
get 0 modulo 2 in the first coordinate, and element in the annihilator must be an even
integer. Then, we are done as the second coordinate becomes 0 modulo 6. Hence

I = 4Z 6= Ann(Ann(I)) = 2Z.

Exercise 18. Let F = R, let V = R2 and let T be the linear transformation from V to V
which is rotation clockwise about the origin by π/2 radians. Show that V and 0 are the only
F [x]-submodules for this T .
Proof. Analogously, we can show that the only T stable subspaces of R2 are 0 and R2 . Let
N be a T stable subspace and let n be its dimension. Since we are in R2 it suffices to show
that if n = 1, then the space is not stable. Just note that in R2 orthogonal vectors are
linearly independant. Therefore. if v is the basis vector of a 1 dimensional space N , T (v) is
independant from v and hence is not in T (N ). The conclusion then follows as ) and R2 are the
trivial invariant subspaces.

Exercise 20. Let F = R, let V = R2 and let T be the linear transformation from V to V
which is rotation clockwise about the origin by π radians. Show that every subspace of V is an
F [x]-submodule for this T .
Proof. As in the previous exercise, it suffices to show that every subspace of R2 is T invariant.
Since 0 and R2 are trivially invariant we only need to show that any 1-dimensional space is T
invariant. Let {v} be the basis of the proposed subspace (note v 6= 0). Then any vector r · v in
the subspace satisfies
T (r · v) = r · T (v) = r · (−v) = (−r) · v.
It is clear that T (v) is in the same subspace, as we wanted.

2
Section 10.2

Exercise 5. Exhibit all Z-module homomorphisms from Z/30Z to Z/21Z.

Exhibition. Note that Z/30Z is generated by 1, and furthermore, over the integers we have
ϕ(rx + y) = rxϕ(1) + yϕ(1).
Therefore, it suffices to define the map 1 7→ ϕ(1) to characterize the homomorphism. We
proceed by looking at the order of 1. It is necessary that 0 = 30ϕ(1), therefore we look for
solutions to
30ϕ(1) ≡ 0 mod 21.
Since 21 = 3 cot 7, ϕ(1) must be a multiple of 7. Then, the possible choices are 7, 14, and 21
and the homomorphisms are characterized by
ϕ(1) = 7
ϕ(1) = 14
ϕ(1) = 21 = 0.

Exercise 8. Let ϕ : M → N be an R-module homomorphism. Prove that ϕ(Tor(M )) ⊆

Tor(N ).
Proof. Let n be an arbitrary element of ϕ(Tor(M )). Then, there exists m ∈ Tor(M ) such that
n = ϕ(m). Furthermore, there exist r ∈ R such that rm = 0. Then, we see that
0 = ϕ(0) = ϕ(rm) = rϕ(m) = rn.
Therefore n ∈ Tor(N ) as desired.

Exercise 13. Let I be a nilpotent ideal in a commutative ring R, let M and N be R-modules
 and
letϕ : M → N be an R-module homomorphism. Show that if the induced map ϕ̄ : M IM →
N IN is surjective, then ϕ is surjective.
Proof. Observe that the map ϕ̄ is given by (m + IM ) 7→ (φ(m) + IN ). Then
  
N IN = ϕ̄(M IM ) = (ϕ(M ) + IN ) IN.
Then, by the lattice isomorphism theorem
N = ϕ(M ) + IN.
Before we proceed, we prove the following,
Iϕ(M ) ⊆ ϕ(M ).
Take n ∈ Iϕ(M ) and note that n has the form iϕ(m) for some i ∈ I and m ∈ M . Then
n = iϕ(m) = ϕ(im), and it follows that n ∈ ϕ(M ). Back to the equality N = ϕ(M ) + IN., we
plugin recursively to see that
N = ϕ(M ) + IN
= ϕ(M ) + I(ϕ(M ) + IN )
= ϕ(M ) + Iϕ(M ) + I 2 N
= ϕ(M ) + I 2 N since Iϕ(M ) ⊆ ϕ(M ).

3
Inductively we see that N = ϕ(M ) + I n N . In particular, since I is nilpotent, of say, order n.

N = ϕ(M ) + I n N = ϕ(M ) + 0N = ϕ(M ).

Hence, ϕ is surjective as we wanted.

Section 10.3
Exercise 5. Let R be an integral domain. Prove that every finitely generated torsion R-
module has a nonzero annihilator i.e., there is a nonzero element r ∈ R such that rm = 0 for
all m ∈ M — here r does not depend on m. Give an example of a torsion R-module whose
annihilator is the zero ideal.
Proof. Let M be the torsion R − module. Then, any m ∈ M has the form

m = r1 g1 + r2 g2 + · · · + rn gn ,

where ri ∈ R, gi is a generator, and n is finite as M is finitely generated. Let ai be a nonzero

element such that,Q ai gi = 0 (this element exist by definition of a torsion module) and consider
the product p = ni=1 ai . Observe that p is non zero since R is an integral domain and ai 6= 0
for all i. Then, for an arbitrary element m of the torsion module M we have,
n6=i
n
!
X Y
pm = ri ai gi ai = 0.
i=1

Thus, p is a non zero element in AnnR M , as desired.

Exercise 9. An R-module M is called irreducible if M 6= 0 and if 0 and M are the only
submodules of M . Show that M is irreducible if and only if M 6= 0 and M is a cyclic module
with any nonzero element as generator. Determine all the irreducible Z-modules.
Proof. First, assume M is irreducible. Then, by definition, M is not 0. Now, let m be any
nonzero element in M . Since 1 ∈ R then 1 · m = m is in Rṁ. The latter forces R · m = M .
Hence M is cyclic and its generated by any nonzero element.
Now we assume that M 6= 0 and M is cyclic, generated by any non zero element. It suffices to
show that 0 and M are the only submodules. Note that the additive identity of M generates
the 0 submodule. On the other hand, if N is a nonzero submodule of M then N contains a
nonzero element n. Furthermore R · n ⊆ N is a subset of M . Since M is generated by any
nonzero element we have R · n = M ⊆ N . Hence N = M as desired.
It follows that the irreducible Z-modules are the cyclic groups of prime order.
Exercise 10. Assume R is commutative. Show that an R-module M is irreducible if and only
if M is isomorphic (as an R-module) to R I where I is a maximal ideal of R.[By the previous
exercise, if M is irreducible there is a natural map R → M defined by r 7→ rm, where m is any
fixed nonzero element of M .]

Proof. First we assume that M is isomorphic to R I. Note that, since I is maximal, the
quotient R I is a field, with trivial ideals. Now, since a submodule of M is one to one corre-
spondence with ideals in R I (by the fourth isomorphism theorem), the only submodules of M
are 0 and M itself. Hence M is irreducible.
Now we assume that M is irreducible. Taking the hint, we let ϕ : R → M  be given by
ϕ(r) = rm, for some fixed m 6= 0 in M . By the first isomorphism theorem R ker(ϕ) is iso-
morphic to M . It is left to show that ker(ϕ) is a maximal ideal in R. This is easy, as ideals

4
 
in R ker(ϕ) are in one to one correspondence with submodules of M (again, by the fourth
isomorphism theorem). Since the only submodules of M are0 and M , it follows that the only
ideals in R ker(ϕ) are 0 and the entire quotient. Thus, R ker(ϕ) is a field and ker(ϕ) is a
maximal ideal as we wanted.
Exercise 11. Show that if M1 and M2 are irreducible R-modules , then any nonzero R-
module homomorphism from M1 to M2 is an isomorphism. Deduce that if M is irreducible then
EndR (M ) is division ring.[Consider the kernel and the image.]
Proof. Let ϕ be a nonzero homomorphism from M1 to M2 . First we show that ϕ is injective
by proving that ker(ϕ) = {0}. By way of contradiction, assume there exist m1 6= 0 in ker(ϕ).
Then, since M1 is irreducible and furthermore generated by m1 , for all m ∈ M1 there exist
r ∈ R such that rm1 = m. Then,

ϕ(m) = ϕ(rm1 ) = rϕ(m1 ) = 0.

It follows that ϕ(m) is 0 for all m ∈ M , a contradiction since ϕ is not a zero homomorphism.
Now we show that ϕ is surjective. Since ϕ is not the zero homomorphism, there exist m2 6= 0
such that m2 = ϕ(m1 ). Furthermore since M2 is generated by m2 , we have that, for all m ∈ M2 ,
m = rm2 for some r ∈ R. Thus,

m = rm2 = rϕ(m1 ) = ϕ(rm1 ).

Since m is an arbitrary element in M2 we have shown that m is in the image of ϕ for all m ∈ M2 .
Hence, ϕ being injective and surjective implies it is an isomorphism.
Now consider the ring EndR (M ), where M is irreducible. We have shown that any nonzero ϕ
in the ring is an isomorphism, hence invertible. Therefore, EndR (M ) satisfies the definition of
a division ring.