Chapter 1

Algebraic Sets

1.1 Affine Space

In elementary geometry, one considered figures with coordinates in some Carte-
sian power of the real numbers. As our starting point in algebraic geometry,
we will consider figures with coordinates in the Cartesian power of some fixed
field k.

1.1.1 Definition. Let k be a field, and let An (k) = {(a1 , . . . , an ) | a1 , . . . , an ∈

k}. When the field is clear, we will shorten An (k) to An . We will refer to An
as affine n-space. In particular, A1 is called the affine line, and A2 is called
the affine plane.

From the algebraic point of view, the most natural functions to consider
on An are those defined by evaluating a polynomial in k[x1 , . . . , xn ] at a point.
Analogously, the simplest geometric figures in An are the zero sets of a single
polynomial.

1.1.2 Definition. If f ∈ k[x1 , . . . , xn ], a point p = (a1 , . . . , an ) ∈ An such

that f (p) = f (a1 , . . . , an ) = 0 is called a zero of f and

V(f ) = {p ∈ An | f (p) = 0}

is called the zero set or zero locus of f . If f is non-constant, V(f ) is called

the hypersurface defined by f . A hypersurface in An is also called an affine
surface, in order to distinguish it from hypersurfaces in other ambient spaces.

1.1.3 Examples.
(i) In R1 , V(x2 + 1) = ∅, but in C1 , V(x2 + 1) = {±i}. Generally, if n = 1
then V(F ) is the set of roots of F in k. If k is algebraically closed and F
is non-constant then V(F ) is non-empty.
(ii) In Z1p , by Fermat’s Little Theorem, V(xp − x) = Z1p .
(iii) By Fermat’s Last Theorem, if n > 2 then V(xn + y n − 1) is finite in Q2 .

1
 (iv) In R2 , V(x2 +y 2 −1) = the unit circle in R2 , and in Q2 it gives the rational
points on the unit circle. Notice the circle admits a parameterization by
rational functions as follows:
1 − t2
 
2t
(x, y) = , , t ∈ R.
1 + t2 1 + t2

When t ∈ Z then we get a point in Q2 .

Remark. A rational curve is a curve that admits a parameterization by rational

functions. For example, the curve in the last example is rational.

1.2 Algebraic Sets and Ideals

1.2.1 Definition. If S is any set of polynomials in k[x1 , . . . , xn ], we define
\
V(S) = {p ∈ An | f (p) = 0 for all f ∈ S} = V(f )
f ∈S

If S = {f1 , . . . , fn } then we may write V(f1 , . . . , fn ) for V(S). A subset X ⊆ An

is an (affine) algebraic set if X = V(S) for some S ⊆ k[x1 , . . . , xn ]

1.2.2 Examples.
(i) For any a, b ∈ k, {(a, b)} is an algebraic set in k2 since {(a, b)} = V(x −
a, y − b).
(ii) In R2 , V(y−x2 , x−y 2 ) is only 2 points, but in C2 it is 4 points. Generally,
Bézout’s Theorem tells us that the number of intersection points of a
curve of degree m with a curve of degree n is mn in projective space over
an algebraically closed field.
(iii) The twisted cubic is the rational curve {(t, t2 , t3 ) | t ∈ R} ⊆ R3 . It is an
algebraic curve; indeed, it is easy to verify that it is V(y − x2 , z − x3 ).
(iv) Not all curves in R2 are algebraic. For example, let

X = {(x, y) | y − sin x = 0}

and suppose that X is algebraic, so that X = V(S) for some S ⊆ R[x, y].
Then there is F ∈ S such that F 6= 0 and so
\
X = V(S) = V(f ) ⊆ V(F ).
f ∈S

Notice that the intersection of X with any horizontal line y − c = 0 is

infinite for −1 ≤ c ≤ 1. Choose c such that F (x, c) is not the zero
polynomial and notice that the number of solutions to F (x, c) = 0 is
finite, so X cannot be algebraic.

2
Remark. The method used in the last example works in more generality. Sup-
pose that C is an algebraic affine plane curve and L is a line not contained C.
Then L ∩ C is either ∅ or a finite set of points.

1.2.3 Proposition. The algebraic sets in A1 are ∅, finite subsets of A1 , and

A1 itself.

Proof: Clearly these sets are all algebraic. Conversely, the zero set of any
non-zero polynomial is finite, so if S contains a non-zero polynomial F then
V(S) ⊆ V(F ) is finite. If S = ∅ or S = {0} then V(S) = A1 . 

Before we continue, we recall some notation. If R is a ring and S ⊆ R, then

hSi denotes the ideal generated by S 1 . If S = {s1 , . . . , sn }, then we denote hSi
by hs1 , . . . , sn i.
1.2.4 Proposition.
(i) If S ⊆ T ⊆ k[x1 , . . . , xn ] then V(T ) ⊆ V(S).
(ii) If S ⊆ k[x1 , . . . , xn ] then V(S) = V(hSi), so every algebraic set is equal
to V(I) for some ideal I.

Proof:
(i) Since S ⊆ T ,
\ \
V(T ) = V(f ) ⊆ V(f ) = V(S).
f ∈T f ∈S

(ii) From (i), V(hSi) ⊆ V(S). If x ∈ V(S) and f ∈ I then we can write f as

f = gq f1 + · · · + gm fm ,

where fi ∈ S and gi ∈ k[x1 , . . . , xn ]. Then

f (x) = g1 (x)f1 (x) + · · · + gm (x)fm (x) = 0

since x ∈ V(S). 

Since every algebraic set is the zero set of an ideal of polynomials, we now
turn our attention to ideals in polynomial rings. If a ring R is such that all of
its ideals are finitely generated it is said to be Noetherian 2 . For example, all
fields are Noetherian. The Hilbert Basis Theorem states that all polynomial
rings with coefficients in a Noetherian ring are Noetherian.
1 The ideal generated by S is the intersection of all ideals containing S. More concretely,
( n )
X
hSi = ak sk : a1 , . . . , an ∈ R and s1 , . . . , sn ∈ S .
k=1

2 Some readers may be more familiar with a different definition of Noetherian in terms of
ascending chains of ideals. This definition is equivalent to ours by Proposition A.0.9.

3
1.2.5 Theorem (Hilbert Basis Theorem). If R is Noetherian, then R[x1 , . . . , xn ]
is Noetherian.

Proof: See Appendix A. 

An important geometric consequence of the Hilbert Basis Theorem is that every

algebraic set is the zero set of a finite set of polynomials.

1.2.6 Corollary. Every algebraic set X in An is the zero set of a finite set of
polynomials.

Proof: k[x1 , . . . , xn ] is Noetherian, so if X = V (S), then X = V (hSi) =

V (S 0 ), where S 0 is a finite subset of k[x1 , . . . , xn ] that generates hSi. 

Tn
Remark. Since V(f1 , . . . , fn ) = k=1 V(fk ), the preceding corollary shows that
every algebraic set is the intersection of finitely many hypersurfaces.

1.2.7 Proposition.
S T
(i) If {Iα } is a collection of ideals then V( α Iα ) = α V(Iα ), so the inter-
section of any collection of algebraic sets is an algebraic set.
(ii) If I and J are ideals then V(IJ) = V (I) ∪ V(J), so the finite union of
algebraic sets is an algebraic set.3
(iii) V(0) = An , V(1) = ∅, and V(x1 − a1 , . . . , xn − an ) = {(a1 , . . . , an )}, so
any finite set of points is algebraic.

Proof:
(i) We have
!
[ \ \ \ \
V Iα = V(f ) = V(f ) = V(Iα ).
α f ∈∪α Iα α f ∈Iα α

3 Recallthat the product of I and J is the ideal generated by products of an element

from I and an element from J. More concretely,

n
( )
X
IJ = ak bk : a1 , . . . , an ∈ I and b1 , . . . , bn ∈ J .
k=1

4
 (ii) Since (gh)(x) = 0 if and only if g(x) = 0 or h(x) = 0,
\
V(IJ) = V(f )
f ∈IJ
\
= V(gh)
g∈I,h∈J
\
= V(g) ∪ V(h)
g∈I,h∈J
\ \
= V(g) ∪ V(h)
g∈I h∈J

= V(I) ∪ V(J).
(iii) This is clear. 

Remark. Note that finiteness of the union in property (ii) is required; for ex-
ample, consider Z in R. It is not an algebraic set, because a polynomial over a
field can only have finitely many roots, but it is the union of (infinitely many)
algebraic sets, namely V(x − n) for n ∈ Z.

The properties in Proposition 1.2.7 allow us to define a topology4 on An

whose closed sets are precisely the algebraic sets.
1.2.8 Definition. The topology on An whose closed sets are precisely the
algebraic sets is called the Zariski topology.

Remark. When k is one of Q, R, or C, the Zariski topology is weaker than the

usual metric topology, as polynomial functions are continuous, so their zero
sets are closed. However, in each of these cases, the Zariski topology is strictly
weaker than the metric topology. For example, Z is closed in the usual topology
of each of Q, R, or C, but is not algebraic and thus is not closed in the Zariski
topology.

1.2.9 Example. The non-empty open sets in the Zariski topology on the affine
line A1 are precisely the complements of finite sets of points. However, this is
not true for An when k is infinite and n > 1. For example, V(x2 + y 2 − 1), the
unit circle in R2 , is closed but is not finite. Moreover, note that the Zariski
topology on An is Hausdorff5 if and only if k is finite, in which case it is identical
to the discrete topology.
4 A topology on a set X is a collection τ of subsets of X that satisfies the following

properties:
(i) ∅, X ∈ τ , S
(ii) if Gi ∈ τ for every i ∈ I then i∈I Gi ∈ τ ,
(iii) if G1 , G2 ∈ τ then G1 ∩ G2 ∈ τ .
The sets in τ are said to be open, and their complements are said to be closed.
5 Recall that a topology is said to be Hausdorff if distinct points always have disjoint

open neighbourhoods.

5
 We have associated an algebraic subset of An to any ideal in k[x1 , . . . , xn ]
by taking the common zeros of its members. We would now like to do the
converse and associate an ideal in k[x1 , . . . , xn ] to any subset of An .

1.2.10 Definition. Given any subset X ⊆ An we define I(X) to be the ideal

of X,
I(X) = {f ∈ k[x1 , . . . , xn ] | f (p) = 0 for all p ∈ X}.

1.2.11 Examples.
(i) The following ideals of k[x] correspond to the algebraic sets of A1 : I(∅) =
h1i, I({a1 , . . . , an }) = h(x − a1 ) · · · (x − an )i, and
(
1 0 if k is infinite,
I(A ) = pn
hx − xi if k has pn elements.

Note that if X ⊆ A1 is infinite then k is infinite and I(X) = 0.

(ii) In A2 , I({(a, b)}) = hx − a, y − bi. Clearly

hx − a, y − bi ⊆ I({(a, b)}),

so we need only prove the reverse inequality. Assume that f ∈ I({(a, b)}).
By the division algorithm, there is g(x, y) ∈ k[x, y] and r(y) ∈ k[y] such
that
f (x, y) = (x − a)g(x, y) + r(y).
But 0 = f (a, b) = r(b), so y − b divides r(y) and we can write we can
write r(y) = (y − b)h(y), and hence

f = (x − a)g + (y − b)h ∈ hx − a, y − bi.

1.2.12 Proposition.
(i) If X ⊆ Y ⊆ An then I(Y ) ⊆ I(X).
(ii) I(∅) = k[x1 , . . . , xn ].
I({(a1 , . . . , an )}) = hx1 −a1 , . . . , xn −an i for any point (a1 , . . . , an ) ∈ An .
I(An ) = 0 if k is infinite.
(iii) S ⊆ I(V(S)) for any set of polynomials S ⊆ k[x1 , . . . , xn ].
X ⊆ V(I(X)) for any set of points X ⊆ An .
(iv) V(I(V(S))) = V(S) for any set of polynomials S ⊆ k[x1 , . . . , xn ].
I(V(I(X))) = I(X) for any set of points X ⊆ An .

Proof:
(i) If f is zero on every point of Y then it is certainly zero on every point of
X.

6
 (ii) That I(∅) = k[x1 , . . . , xn ] and I(An ) = 0 if k is infinite are clear. Fix
(a1 , . . . , an ) ∈ An , and define ϕ : k[x1 , . . . , xn ] → k by ϕ(f ) = f (a1 , . . . , an ).
Clearly, ϕ is a surjective homomorphism, and

ker(ϕ) = hx1 − a1 , . . . , xn − an i.

We have
k[x1 , . . . , xn ]/hx1 − a1 , . . . , xn − an i ∼
= k,
so hx1 − a1 , . . . , xn − an i is a maximal ideal. The ideal I({(a1 , . . . , an )})
is proper and contains hx1 − a1 , . . . , xn − an i, a maximal ideal, so it must
be equal to that maximal ideal.
(iii) These follow from the definitions of I and V.
(iv) From (iii), V(S) ⊆ V(I(V(S))), and by Proposition 1.2.4 (i), V(I(V(S))) ⊆
V(S) since S ⊆ V(I(S)). Therefore V(S) = V(I(V(S))). The proof of
the second part is similar. 

Remarks.
(i) As is shown in the proof of part (ii) of the last proposition, the ideal
hx1 − a1 , . . . , xn − an i of any point (a1 , ..., an ) ∈ An is maximal.
(ii) Equality does not always hold in part (iii) of the last proposition, as
shown by the following examples:
(a) Consider I = hx2 + 1i ⊆ R[x]. Then 1 ∈ / I, so I 6= R[x]. But
V(I) = ∅, so I(V(I)) = R[x] % I.
(b) Consider X = [0, 1] ⊆ R. Then I(X) = 0 and V(I(X)) = R % X.
These examples also show that not every ideal of k[x1 , . . . , xn ] is the ideal
of a set of points and that not every subset of An is algebraic.

We have a correspondence between subsets of An and ideals of k[x1 , . . . , xn ]

given by
X 7→ I(X) and I 7→ V (I).
By part (iv) of the last proposition, this correspondence is one-to-one when
restricted to algebraic sets and ideals of sets of points. Given that not every
subset of An is algebraic and not every ideal of k[x1 , . . . , xn ] is the ideal of a
set of points, we would like to examine the smallest algebraic set containing
an arbitary subset of An and the smallest ideal of a set of points containing an
arbitrary ideal of k[x1 , . . . , xn ].
1.2.13 Definition. Let X ⊆ An and I ⊆ k[x1 , . . . , xn ] be an ideal. The
closure of X (in the Zariski topology) is the smallest algebraic set containing
X (i.e. the smallest closed set containing X), and is denoted X. The closure
of I is the smallest ideal of a set of points that contains I, and is denoted I. If
I = I, we say that I is closed .

Remark. Note that I is the ideal of a set of points if and only if I = I.

7
1.2.14 Proposition.
(i) If X ⊆ An , then X = V(I(X)).
(ii) If I ⊆ k[x1 , . . . , xn ] is an ideal, then I = I(V(I)).

Proof: We will only prove (i), as the proof of (ii) is very similar. By part
(iii) of Proposition 1.2.12, we have X ⊆ V(I(X)) Since V(I(X)) is an algebraic
set, X ⊆ V(I(X)). Conversely, since X ⊆ X, V(I(X)) ⊆ V(I(X)). By part
(ii) of Proposition 1.2.7, we have V(I(X)) = X, because X is an algebraic set.
Therefore, V(I(X)) ⊆ X, showing that X = V(I(X)). 

1.2.15 Examples.
(i) If X = (0, 1) ⊆ R, then the closure of X in the metric topology is [0, 1],
whereas the closure of X in the Zariski topology is R.
(ii) If k is infinite and X ⊆ A1 is any infinite set of points then X = A1 .
In particular, the Zariski closure of any non-empty open set is the whole
line, or every non-empty open set is Zariski dense in the affine line.
(iii) Let I = hx2 i. Then I = I(V(I)) = hxi, so that I 6= I and I is not an
ideal of a set of points.

1.3 Radical Ideals and the Nullstellensatz

In the previous section, we examined algebraic sets and ideals of sets of points.
We saw that every algebraic set is the zero set of a finite set of polynomials.
In this section, we will look for an intrinsic description of ideals of sets of
points. We have already seen that not every ideal is the ideal of a set of points.
Intuitively, an ideal I of k[x1 , . . . , xn ] is the ideal of a set of points whenever its
generators intersect with the smallest possible multiplicity. However, since the
multiplicity of any intersection is lost when we take the zero set of an ideal, as
sets do not have any way of keeping track of multiplicity, we should not expect
to get it back when we again take the ideal of that zero set.
1.3.1 Examples.
(i) Let I = hx2 +y 2 −1, xi ⊆ R[x, y]. The set V(x2 +y 2 −1) is the unit circle,
and V(x) is the vertical line through the origin. The line intersects the
circle twice, each time with “multiplicity one”. Therefore, our intuition
would lead us to think that I is a closed ideal. This is correct, as

I = I(V(I)) = I({(0, −1), (0, 1)}) = hx, y 2 − 1i = I.

(ii) Let I = hx2 + y 2 − 1, x − 1i ⊆ R[x, y]. The set V(x2 + y 2 − 1) is the unit
circle, and V(x − 1) is the vertical line that is tangent to the circle at
(1, 0). Because it only intersects the circle at one point, the intersection is
with “multiplicity two”. Therefore, our intuition would lead us to think
that I is not a closed ideal. This is indeed the case, as

I = I(V(I)) = I({(1, 0)}) = hx − 1, yi =

6 I.

8
 The zero sets of the generators of I are a vertical line through (1, 0) and a
horizontal line through the origin, which intersect once at the point (1, 0)
with “multiplicity one”, again confirming our intuition.

Algebraically, if I = I(X) for some X ⊆ An then √ I is radical. Recall that

an ideal I is radical if I is equal to its radical ideal I,
√
I = {a ∈ R | an ∈ I for some n > 0}.

Equivalently, I is radical if the following condition holds:

an ∈ I implies that a ∈ I for all a ∈ R and n > 0.

(See Proposition A.0.14.)

1.3.2 Examples.
(i) If X ⊆ An then I(X) is radical, because f (x) = 0 whenever f n (x) = 0.
(ii) Every prime ideal is radical. For a proof, see Proposition A.0.15. How-
ever, not every proper radical ideal is prime. For example, the ideal

hx(x − 1)i = I({0, 1})

of k[x] is radical, but it is not prime.

(iii) Let I = hx2 + y 2 − 1, x − 1i ⊆ R[x, y]. Then y 2 ∈ I, because

y 2 = (x2 + y 2 − 1) − (x + 1)(x − 1),

but y ∈
/ I, simply because of the degrees of the y terms in the generators.
Hence I is not radical. We already examined this example geometrically
above.
(iv) Let I = hy − x2 , y − x3 i. If u = x(x − 1), then

u2 = [(y − x2 ) − (y − x3 )](x − 1) ∈ I,

but u ∈/ I, because of the degrees of the x terms in the generators. Hence I

is not radical. Geometrically, V(y − x2 ) is an upwards parabola through
the origin, and V(y − x3 ) intersects it twice, at the origin and at the
point (1, 1). There are only two points of intersection, yet the degrees
of the polynomials involved imply that there should be three, including
multiplicity. Thus one of the points of intersection (in fact, the origin)
has “multiplicity two”.

We saw in the first of the above examples that if I is the ideal of a set of
points then I is radical. Is the converse true? That is, if I is radical is it true
that I = I?
√
1.3.3 Proposition. If I is an ideal of k[x1 , . . . , xn ], then I ⊆ I ⊆ I. In
particular, a closed ideal is radical.

9
 √ √
Proof: Clearly, I ⊆ I. Suppose f ∈ I. Then f n ∈ I for some n ≥ 1. Since
f n (x) = 0 if and only if f (x) = 0, √
we have f ∈ I(V(I)). By Proposition 1.2.14,
I = I(V(I)), so f ∈ I. Therefore, I ⊆ I. 
√
It follows
√ from the previous proposition that if I = I then I = I if and
only if
√ I = I(V(I)). However, if k is not algebraically closed, it often happens
that I 6= I(V(I)):
1.3.4 Example. The polynomial x2 + 1 ∈ R[x] is irreducible, so the ideal
hx2 + 1i is maximal. Hence it is radical, and it is obviously proper. However,

I(V(x2 + 1)) = I(∅) = k[x]

so hx2 + 1i is not an ideal of a set of points. Clearly, x2 + 1 can be replaced by

any irreducible polynomial of degree at least 2 in any non-algebraically closed
field.

However, the lack of algebraic closure in the base field is actually √

necessary
for a counterexample. If the base field is algebraically closed, I = I. This
result is due to Hilbert and is known as the Nullstellensatz, which is German
for “zero points theorem”.
1.3.5 Theorem (Nullstellensatz). Suppose√k is algebraically √ closed, and let
I ⊆ k[x1 , . . . , xn ] be an ideal. Then √
I(V(I)) = I, so I = I and I is the ideal
of a set of points if and only if I = I.

Proof: See Appendix C. 

A related question is the characterization of maximal ideals of k[x1 , . . . , xn ].

We have seen that the ideal of a single point (a1 , . . . , an ) ∈ An is the maximal
ideal hx1 − a1 , . . . , xn − an i. Are all maximal ideals of k[x1 , . . . , xn ] of this
form? Again, the example of hx2 + 1i in R[x] shows this to be false in general.
However, this is true when k is algebraically closed. Indeed, if I is a maximal
ideal of k[x1 , . . . , xn ] then I is radical, so by the Nullstellensatz I is the ideal
of a set of points. Since I is a maximal ideal and taking zero sets reverses
inclusions, V(I) is a non-empty minimal algebraic set, which must consist of a
single point (a1 , . . . , an ) ∈ An .

1.4 Irreducible Algebraic Sets

1.4.1 Definition. An algebraic set X ⊆ An is irreducible if X 6= ∅ and X
cannot be expressed at X = X1 ∪ X2 , where X1 and X2 are algebraic sets not
equal to X.

1.4.2 Proposition. An algebraic set X ⊆ An is irreducible if and only if I(X)

is prime.

10
Proof: If X is irreducible then suppose that f, g ∈ k[x1 , . . . , xn ] are such that
f g ∈ I(X). Then hf gi ⊆ I(X), so X = V(I(X)) ⊆ V(f g) = V(f )∪V(g). Hence
X = (X ∩ V(f )) ∪ (X ∩ V(g)), so without loss of generality, X = X ∩ V(f ) ⊆
V(f ). Therefore f ∈ I(X) and I(X) is prime.
Suppose that I(X) is prime but is reducible, with X = X1 ∪ X2 . Then
I(X) = I(X1 ) ∩ I(X2 ). If I(X) = I(X1 ) then X = X1 , which is not allowed.
Hence there is f ∈ I(X1 ) \ I(X). But for any g ∈ I(X2 ), f g ∈ I(X1 ) ∩ I(X2 ) =
I(X), so since f ∈ / I(X) and I(X) is prime, g ∈ I(X). This implies that
I(X) = I(X2 ) (and hence X = X2 ), a contradiction. 

1.4.3 Examples.
(i) An is irreducible for all n ≥ 1, because I(An ) = {0}, which is a prime
ideal.
(ii) If (a1 , . . . , an ) ∈ An , then {x} is irreducible, because

I({(a1 , . . . , an )}) = hx1 − a1 , . . . , xn − an i,

which is a maximal ideal and therefore prime.

(iii) Since k[x1 , . . . , xn ] is a UFD, any ideal generated by an irreducible poly-
nomial is prime. If k is algebraically closed then V(p) is irreducible
for every irreducible polynomial p ∈ k[x1 , . . . , xn ] by the Nullstellensatz.
Hence when k is algebraically closed there is a one to one correspondence
between irreducible polynomials in k[x1 , . . . , xn ] and irreducible hyper-
surfaces in An .

Remark. If X ⊆ An is an irreducible algebraic set, then X is connected in the

Zariski topology. Recall that a closed subset of a topological space is connected
if and only if it is not the union of two disjoint closed proper subsets. However,
if X = X1 ∪ X2 where X1 , X2 ⊆ An are closed, X = X1 or X = X2 by the
irreducibility of X, showing that X is connected.

The correspondence between algebraic sets and ideals of sets of points takes
irreducible algebraic sets to prime ideals, and prime ideals that are ideals of sets
of points to irreducible algebraic sets. If k is algebraically closed, by combining
the results of this chapter we have the following correspondence:

Geometry Algebra
affine space An polynomial ring k[x1 , . . . , xn ]
algebraic set radical ideal
irreducible algebraic set prime ideal
point maximal ideal

Remark. If k is not algebraically closed then there are more prime ideals than
irreducible algebraic sets.

11
 (i) distinct prime ideals may give the same algebraic set, e.g. V(hx2 + y 2 i) =
{(0, 0)} = V(hx, yi) in R2 ;
(ii) a prime ideal may have a reducible zero set, e.g. V(hx2 + y 2 (y − 1)2 i) =
{(0, 0), (0, 1)} in R2 .

Mirroring the decomposition of an integer as the product of primes, every

algebraic set decomposes as the union of finitely many irreducible algebraic
sets.

1.4.4 Proposition. Every algebraic set X is a finite union of irreducible al-

gebraic sets.

Proof: Suppose that X is not the union of a finite number of irreducibles.

Then, in particular, X itself is not irreducible, so X = X1 ∪X10 , where X1 , X10 $
X. Without loss of generality, we can assume that X1 is not the union of a finite
number of irreducibles. Repeating this we get an infinite strictly descending
chain of algebraic sets X % X1 % · · · . But then I(X) $ I(X1 ) $ · · · is an
infinite strictly ascending chain of ideals in k[x1 , . . . , xn ], a contradiction since
k[x1 , . . . , xn ] is Noetherian. 

Suppose that X = X1 ∪ · · · ∪ Xr , where each Xi is an irreducible algebraic

set. In what sense is this decomposition unique? It can not literally be unique,
as we could include any irreducible algebraic subset of X. However, this is the
only obstruction to the uniqueness of the decomposition, since any irreducible
algebraic subset of X must in fact already be contained in some Xj , as implied
by the following lemma.

1.4.5 Lemma. Let X ⊆ An be an irreducible algebraic set. If X ⊆ X1 ∪ · · · ∪

Xr , where X1 , . . . , Xr ⊆ An are algebraic, then X ⊆ Xj for some j.
Sn Sn
Proof: Since X ⊆ i=1 Xi , X = i=1 X ∩ Xi . By the irreducibility of X, we
have X = X ∩ Xj for some j, so X ⊆ Xj . 

By successively discarding the Xi ’s that are included in one of the other

Xj ’s, we therefore obtain a description of X as

X = X1 ∪ · · · ∪ Xm ,

where each Xi is an irreducible algebraic set and Xi ( Xj when i 6= j. We call

such a decomposition an irredundant decomposition of X. Since the following
proposition shows that an algebraic set has a unique irredundant decomposi-
tion, we will usually refer to an irredundant decomposition of X simply as the
decomposition of X.

1.4.6 Proposition. Every algebraic set X has a unique irredundant decom-

position into irreducible algebraic sets.

12
Proof: By Proposition 1.4.4, X is the finite union of irreducible algebraic sets.
By possibly removing some constituents of this union, we have an irredundant
decomposition X = X1 ∪ · · · ∪ Xm . Suppose that X also has an irredundant
decompostion X = Y1 ∪ · · · ∪ Yn . Then for any i, Xi is contained in some
Yj0 by Lemma 1.4.5. Similarly, Yj0 ⊆ Xi0 for some i0 , but this implies that
Xi ⊆ Yj0 ⊆ Xi0 , and since the decomposition is irredundant, Xi = Xi0 = Yj0 .
Therefore every Xi corresponds to a Yj , and vice-versa. 

1.4.7 Examples.
(i) Suppose that f ∈ k[x1 , . . . , xn ] and f = f1r1 . . . fm
rm
then
V(f ) = V(f1 ) ∪ · · · ∪ V(fm ).
If k is algebraically closed then this is a decomposition and I(V(f )) =
hf1 · · · fm i.
(ii) Consider X = V(y 4 − x2 , y 4 − x2 y 2 + xy 2 − x3 ) ⊆ C2 . Notice that
y 4 − x2 = (y 2 − x)(y 2 + x),
and
y 4 − x2 y 2 + xy 2 − x3 = (y 2 + x)(y − x)(y + x),
so V(y 2 + x) is an irreducible component of X. The other 3 points in X
are (0, 0), (1, 1) and (1, −1). But (0, 0) ∈ V(y 2 + x), so the decomposition
of X is V(y 2 + x) ∪ {(1, 1)} ∪ {(1, −1)}.
(iii) Consider X = V(x2 + y 2 (y − 1)2 ) ⊆ R2 . X = {(0, 0), (0, 1)}, so X is
reducible. But f (x, y) = x2 + y 2 (y − 1)2 is irreducible in R[x, y]. Indeed,
f (x, y) = (x + iy(y − 1))(x − iy(y − 1)).
Since R[x, y] ⊆ C[x, y] are UFDs, if f factors in R[x, y] the decompostion
must agree with the decomposition we have, up to constant multiple, but
this is impossible.

1.5 Classification of Irreducible Algebraic Sets in A2

While the irreducible algebraic subsets of A1 (k) are easy to determine, this is
not the case for An (k) in general. Nevertheless, such a classification exists for
A2 (k). If k is finite then so is A2 (k), so the irreducible algebraic subsets of
A2 (k) are precisely the singletons. Therefore, we assume that k is infinite for
the remainder of this section.
There are only a few possible candidates for irreducible subsets of A2 . Since
k is infinite, A2 itself is irreducible, and any singleton is irreducible. More-
over, it is natural to consider the zero set V(f ) of an irreducible polynomial
f ∈ k[x, y]. However, if V(f ) consists of a finite set of points other than a
singleton, then V(f ) is reducible. But we will show that if V(f ) is infinite it is
always irreducible, and that the sets listed are precisely the irreducible algebraic
subsets of A2 . First, we will prove a proposition that is also of independent
interest.

13
1.5.1 Proposition. If f, g ∈ k[x, y] have no common factors then V(f, g) =
V(f ) ∩ V(g) is at most a finite set of points.

Proof: Since f and g have no common factor in k[x, y] = k[x][y], they have
no common factors in k(x)[y]. Therefore gcd(f, g) exists and is 1 in k(x)[y], so
there are s, t ∈ k(x)[y] such that sf + tg = 1. Hence there is d ∈ k[x] such that
ds = a, dt = b, where a, b ∈ k[x][y] = k[x, y]. Then af + bg = d ∈ k[x]. Now if
(x0 , y0 ) ∈ V(f, g) then d(x0 ) = 0, so there are at most finitely many possible
values for x0 . Similarily, there are at most finitely many possible values for y0 ,
so V(f, g) is finite. 

Remark. Proposition 1.5.1 can be viewed as a weak form of Bézout’s Theorem,

which states that the number of intersection points of a curve of degree m with
a curve of degree n is mn in projective space over an algebraically closed field.

1.5.2 Corollary. If f ∈ k[x, y] is irreducible and X is an infinite algebraic

set such that X ⊆ V(f ), then I(X) = hf i. Therefore, X = V (f ) and V(f ) is
irreducible.

Proof: Clearly, hf i ⊆ I(X). Suppose that there is g ∈ I(X) such that g ∈/ hf i.

Then f and g have no common factors, so V(f, g) is a finite set of points. But
X ⊆ V(f, g) is infinite, so I(X) = hf i and X = V(I(X)) = V(f ). In particular,
if X = V(f ) then I(X) = hf i, which is prime given that f is irreducible, so
V(f ) is irreducible. 

1.5.3 Theorem. Suppose k is infinite. Then the irreducible algebraic sets in

A2 are
(i) A2 ,
(ii) {(a, b)}, for a, b ∈ k,
(iii) V(f ) where f ∈ k[x, y] is irreducible and V(f ) is an infinite set.

Proof: We have already seen that all algebraic subsets of A2 of these forms
are irreducible. Let X ⊆ A2 be an irreducible algebraic set. Assume that X
is not A2 or a single point. Then I(X) 6= 0, so there is at least on non-zero
polynomial f ∈ I(X). Moreover, any irreducible factor of f is in the prime
ideal I(X), since X is assumed to be irreducible. We may therefore assume
that f is irreducible. Then Corollary 1.5.2 implies that X = V (f ) since X is
infinite. 

1.5.4 Examples.
(i) In R2 , V(y − x2 ) is irreducible because f = y − x2 is an irreducible
polynomial and V(y − x2 ) is infinite.
(ii) In R2 , V(y 2 − x2 (x − 1)) is also irreducible for the same reasons. Hence
it is connected in the Zariski topology. However, it is not connected in
the metric topology.

14
Appendix A

Some Ring Theory

A.0.5 Definition. A principal ring is a ring for which every ideal is generated
by a single element. A principal integral domain is called a principal ideal
domain, or PID for short.

A.0.6 Proposition. k[x] is a PID.

Proof: Since k[x] is clearly an integral domain, we only need to show that
it is principal. Let I be an ideal of k[x], and let f be a monic polynomial of
minimum degree in I. First, we show that f is unique, i.e. if g is another monic
polynomial in I such that deg(g) = deg(f ), then f = g. Let h = f − g. Then
h ∈ I, and since deg(h) < deg(f ) we must have h = 0, so g = f .
We now show that I = hf i. Since f ∈ I, we have hf i ⊆ I. To establish the
reverse inclusion, fix g ∈ I. By the division algorithm, there exist q, r ∈ k[x]
such that r is monic, g = qf + r, and either r = 0 or deg(r) < deg(f ). Since I
is an ideal, r = g − qf ∈ I. By the minimality of the degree of f , we can not
have deg(r) < deg(f ), so r = 0. Therefore, g = qf and g ∈ hf i. Since g ∈ I
was arbitrary, this shows that I ⊆ hf i, and thus I = hf i. 

A.0.7 Proposition. If n > 1, k[x1 , . . . , xn ] is not principal.

Proof: Suppose that I is principal. Let I = hx1 , . . . , xn i. Then I = hpi

for some p ∈ k[x1 , . . . , xn ]. Hence p|q for every q ∈ I. In particular, q|xi
for 1 ≤ i ≤ n. Since the only elements in k[x1 , . . . , xn ] that divide every
indeterminate are the non-zero scalars, p must be a scalar. However, this a
contradiction, as there are no non-zero scalars in I. Therefore, our assumption
that I is principal is false, and k[x1 , . . . , xn ] is not principal. 

A.0.8 Definition. We say that a ring R is Noetherian if every ideal of R is

finitely generated.

A.0.9 Proposition. Let R be a ring. Then the following are equivalent:

15
 (i) R is Noetherian,
(ii) R satisfies the ascending chain condition on ideals, i.e. if

I0 ⊆ I1 ⊆ · · · ⊆ In ⊆ · · ·

is a chain of ideals of R, there exists a k ∈ N such that

Ik = Ik+1 = · · · = Ik+n = · · · .

Proof: Suppose R is Noetherian, and let

I0 ⊆ I1 ⊆ · · · ⊆ In ⊆ · · ·

be a chain of ideals of R. Let

[
I= Ik .
k∈N

In general, the union of ideals is not an ideal, but the union of an increasing
chain of ideals can easily be seen to be an ideal. Thus I is an ideal. Since R
is Noetherian, I is finitely generated, i.e. there exist a1 , . . . , am ∈ I such that
I = ha1 , . . . , am i. Let k ∈ N be such that a1 , . . . , am ∈ Ik . Then

I = Ik = Ik+1 = · · · = Ik+n = · · · .

Conversely, suppose R satisfies the ascending chain condition but is not

Noetherian, and let I be an ideal of R that is not finitely generated. Pick
a0 ∈ I, and let I0 = ha0 i. Since I is not finitely generated, I0 6= I. Pick
a1 ∈ I \ I0 , and let I1 = ha0 , a1 i. Since I is not finitely generated, I0 ( I1 6= I.
Continuing by induction, we get an increasing chain of ideals

I0 ( I1 ( · · · ( In ( · · · ,

in contradiction to the ascending condition on R. Therefore, our assumption

that R is not Noetherian is false. 

We now establish that polynomial rings over an arbitrary Noetherian ring

are Noetherian.
A.0.10 Theorem (Hilbert Basis Theorem). If R is a Noetherian ring, then
R[x] is Noetherian.

Proof: Suppose R[x] is not Noetherian, and let I is an ideal of R[x] that is not
finitely generated. Let f0 be a polynomial of minimum degree in I. Continuing
by induction, let fk+1 be a polynomial of minimum degree in I \ hf0 , . . . , fk i.
For every k ∈ N, let dk = deg(fk ), and let ak be the leading coefficient of fk ,
and let J = h{ak : k ∈ N}i. Since R is Noetherian and

ha0 i ⊆ ha0 , a1 i ⊆ · · · ha0 , . . . , an i ⊆ · · ·

16
is an increasing chain of ideals whose union is J, there exists an n ∈ N such
that J = ha0 , . . . , an i.
Let I0 = hf0 , . . . , fn i. By construction, fn+1 ∈
/ I0 . Since J = ha0 , . . . , an i
and an+1 ∈ J, there exist b0 , . . . , bn ∈ R such that an+1 = b0 a0 + · · · + bn an .
Then, as fn+1 ∈ I \ I0 , we have
g = mn+1 − xdn+1 −d0 b0 f0 − · · · − xdn+1 −dn bn fn ∈ I,
so deg(g) < deg(fn+1 ). However, g ∈/ I0 , as fn+1 ∈
/ I0 , contradicting the mini-
mality of deg(fn+1 ). Therefore, our assumption that R[x] is not Noetherian is
false. 

A.0.11 Corollary. If R is a Noetherian ring, then R[x1 , . . . , xn ] is Noetherian.

Proof: Since R[x1 , . . . , xn+1 ] ∼

= R[x1 , . . . , xn ][xn+1 ], the result follows by in-
duction from the Hilbert Basis Theorem. 

A.0.12 Definition. Let R be a ring, and I an ideal in R. The radical of I is

the ideal √
I = {a ∈ R | an ∈ I for some n > 0}.
√
If I = I, we say that I is radical .
√
A.0.13 Proposition. Let R be a ring, and I an ideal of R. Then I is an
ideal of R.
√
Proof: If a ∈ R and b ∈ I, then bn ∈ I for some n > 0, so
(ab)n = an bn ∈ I,
√ √
and ab ∈ I. If a, b ∈ I, am ∈ I and bn ∈ I for some m, n > 0. Therefore,
by the Binomial Theorem,
m+n+1
X m + n − 1
(a + b)m+n+1 = ak bm+n−1−k .
k
k=0

For every k ∈ N, either k ≥ m, or m−1 ≥ k and m+n−1−k ≥ n. This implies

that for any k ∈ N, either ak ∈ I or bm+n−1−k ∈ I. Therefore, every term of
m+n+1
the series expansion
√ of (a +
√ b) is in I, showing that (a + b)m+n+1 ∈ I,
or a + b ∈ I. Therefore, I is an ideal. 

A.0.14 Proposition. Let R be a ring, and I an ideal of R. Then I is radical

if and only if an ∈ I implies that a ∈ I for all a ∈ R and n > 0.
√
Proof: Suppose I is radical and an ∈ I. Then a ∈ I = I. Conversely,
n
√ that a ∈ I implies that a ∈√I for all a ∈ √R and n n> 0. Clearly,
suppose
I ⊆ I, so we only need to show √
that I ⊆ I. If a ∈ I then a ∈ I for some
n > 0. Thus a ∈ I, showing that I ⊆ I and that I is radical. 

17
A.0.15 Proposition. Let R be a ring, and I a prime ideal of R. Then I is
radical.

Proof: Given a ∈ R and n > 0 such that an ∈ I, we will show that a ∈ I by

induction on the n such that an ∈ I. If n = 1 and an ∈ I, then clearly a ∈ I.
Suppose that bn ∈ I implies b ∈ I, and that an+1 ∈ I. Since I is prime, either
a ∈ I or an ∈ I, in which case we also have a ∈ I by our induction hypothesis.
Therefore, by Proposition A.0.14, I is radical. 

18
Appendix B

Transcendence Bases

B.0.16 Definition. Let K be a field, and let F be a subfield of K. A

subset U ⊆ K is said to be algebraically independent over F if for every
n ≥ 1, every non-zero f ∈ F [x1 , . . . , xn ], and all u1 , . . . , un ∈ U , we have
that f (u1 , . . . , un ) 6= 0. A transcendence basis of K over F is an algebraically
independent subset of K that is maximal with respect to inclusion.

B.0.17 Examples.
(i) The empty set is algebraically independent. If K = F , it is also a tran-
scendence basis of K over F .
(ii) Let F a fixed field and let K = F (x1 , . . . , xn ) be the fraction field of the
ring F [x1 , . . . , xn ]. We claim that {x1 , . . . , xn } is a transcendence basis
of K over F . It is clearly algebraically independent, as if f ∈ F [t1 , . . . , tn ]
is such that f (x1 , . . . , xn ) = 0, we have that f = f (t1 , . . . , tn ) = 0. To
show that {x1 , . . . , xn } is a maximal algebraically independent set, we
will show that {x1 , . . . , xn , p/q} is algebraically dependent over F for any
p, q ∈ F [x1 , . . . , xn ], q 6= 0. Define f ∈ F [t1 , . . . , tn+1 ] by

f (t1 , . . . , tn+1 ) = p(t1 , . . . , tn ) − q(t1 , . . . , tn )tn+1 .

Then f 6= 0, but f (x1 , . . . , xn , p/q) = 0, showing that {x1 , . . . , xn , p/q}

is algebraically dependent over F . Therefore, {x1 , . . . , xn } is a transcen-
dence basis of K over F .

B.0.18 Theorem. Let K be a field, and let F be a subfield of K. Then:

(i) Every algebraically independent subset U of K is contained in some tran-

scendence basis. In particular, since the empty set is algebraically inde-
pendent, K has a transcendence basis.
(ii) If B1 and B2 are both transcendence bases of K over F , then card(B1 ) =
card(B2 ).

19
Proof:
(i) Let P be the partial order of algebraically independent subsets S of K that
contain U , ordered by inclusion. If C is a chain in P , then C is clearly
algebraically independent,Sas any possible algebraic dependence involves
finitely many elements of C, which could all be chosen to be in the same
member of C. Therefore, by Zorn’s Lemma, P has a maximal element.
However, by definition, such a maximal element is a transcendence basis
of K containing U .
(ii) For the sake of sanity, we will assume that B1 is finite. In the infinite
case, it is argued using either multiple applications of Zorn’s Lemma or
transfinite induction. Suppose B1 = {x1 , . . . , xm }, where m ≥ 1 is the
minimal cardinality of any transcendence basis. It suffices to show that
if w1 , . . . , wn are algebraically independent elements of K then n ≤ m,
as we could then swap B1 and B2 to get the opposite inequality. If
every wi is an xj , there is nothing to prove, so by possibly reordering the
wi ’s, we can assume that w1 6= xi for i = 1, . . . , m. Since {x1 , . . . , xm }
is a transcendence basis, {w1 , x1 , . . . , xm } is algebraically dependent, so
there is a non-zero polynomial f1 ∈ F [t1 , . . . , tm+1 ], which can clearly be
chosen to be irreducible, such that f1 (w1 , x1 , . . . , xm ) = 0. After possibly
renumbering the xj ’s we may write
k
X
f1 = gj (w1 , x2 , . . . , xm )xj1
j=1

for some k ≥ 1 and g1 , . . . , gk ∈ F [t1 , . . . , tm+1 . No irreducible fac-

tor of gk vanishes on (w1 , x2 , . . . , xn ), otherwise w1 would be a root of
two distinct irreducible polynomials over F (x1 , . . . , xm ). Hence x1 is
algebraic over F (w1 , x2 , . . . , xm ) and w1 , x2 , . . . , xm are algebraically in-
dependent over F , as otherwise the minimality of m would be contra-
dicted. Continuing inductively, suppose that after a suitable renumber-
ing of x1 , . . . , xm we have found w1 , . . . , wr , r < n, sch that K is al-
gebraiic over F (w1 , . . . , wr , xr+1 , . . . , xm ). Then there exists a non-zero
f ∈ F [t1 , . . . , tm+1 ] such that
f (wr+1 , w1 , . . . , wr , xr+1 , . . . , xm ) = 0.
Since the wi ’s are algebraically independent over F , it follows by the
same argument as in the case above that some xj , which we can as-
sume to be xr+1 , is algebraic over F (w1 , . . . , wr+1 , xr+2 , . . . , xm ). Since
a tower of algebraic extensions is algebraic, it follows that K is algebraic
over F (w1 , . . . , wr+1 , xr+2 , . . . , xm ). If n ≥ m, we can continue induc-
tively and replace all of the xj ’s by wi ’s to see that K is algebraic over
F (w1 , . . . , wm ), showing that n = m, as desired. 

B.0.19 Definition. Let K be a field, and let F be a subfield of K. The

transcendence degree of K over F is the cardinality of any transcendence basis
of K over F .

20
Appendix C

A Proof of the Nullstellensatz

We begin by examining extensions of rings that are analogous to algebraic

extensions of fields, where an arbitrary polynomial with coefficients in the base
field is replaced with a monic polynomial with coefficients in the base ring.

C.0.20 Definition. Let S be a ring and R be a subring of S.

(i) An element s ∈ S is integral over R if s is the root of a monic polynomial
in R[x].
(ii) The ring S is an integral extension of R, or just integral over R, if every
s ∈ S is integral over R.
(iii) The integral closure of R in S is the set of elements of S that are integral
over R.

One fundamental fact about integral extensions is that they are transitive,
i.e. the composition of integral extensions is also an integral extension. In the
proof we use a special case of a notion from the theory of modules. If R and
S are rings and R ⊆ S, we say that S is a finitely generated R-module if there
exists a finite set A ⊆ S such that S = RA.

C.0.21 Proposition. Let T be a ring and R be a subring of T . If t ∈ T , then

the following are equivalent:
(i) t is integral over R;
(ii) R[t] is a finitely generated R-module;
(iii) there exists a subring S of T such that t ∈ S and S is a finitely generated
R-module.

Proof: Suppose that t is integral over R. There then exist r0 , ..., rn−1 ∈ R
such that
tn + rn−1 tn−1 + · · · + r1 t + r0 = 0,
or
tn = −(rn−1 tn−1 + · · · + r1 t + r0 ),

21
so tn and all higher powers of t can be expressed as R-linear combinations of
tn−1 , . . . , t, 1. Then R[t] = R{tn−1 , . . . , t, 1} is a finitely generated R-module.
Suppose R[t] is a finitely generated R-module. Since t ∈ R[t] and R[t] ⊆ T ,
this means that S = R[t] satisfies the conditions of (iii).
Suppose there exists a subring S of T such that t ∈ S and S is a finitely
generated R-module. Let A ⊆ S be a finite set such that S = RA. Enumerate
the elements of A by A = {a1 , . . . , an }. For i = 1, . . . , n, the element tai is an
element of S and can thus be written as R-linear combinations of a1 , . . . , an ,
i.e. for some coefficients cij ∈ R,
n
X
tai = cij aj .
j=1

By rearranging terms, we obtain

n
X
(δij t − cij )aj = 0,
j=1

where δij is the Kronecker delta. Let B be the n × n matrix whose i, j entry
is δij t − cij , and let v be the n × 1 column vector whose entries are a1 , . . . , an .
These equations then simply amount to saying that Bv = 0; it follows from
Cramer’s Rule that (det B)ai = 0 for i = 1, . . . , n. Since 1 ∈ S is an R-linear
combination of a1 , . . . , an , it follows that det B = 0. But B = tI − C, where C
is the matrix whose i, j entry is cij . Thus, t is a root of the monic polynomial
det(xI − C) ∈ R[x], i.e. t is integral over R. 

C.0.22 Proposition. Let R, S, and T be rings such that R ⊆ S ⊆ T . If T is

integral over S and S is integral over R, then T is integral over R.

Proof: Fix t ∈ T . Since T is integral over S, there exist s0 , ..., sn−1 ∈ S with

tn + sn−1 tn−1 + · · · + s1 t + s0 = 0.

Since each si ∈ S is integral over R, by Proposition C.0.21, each ring R[si ]

is a finitely generated R-module, implying that R[s0 , . . . , sn−1 ] is a finitely
generated R-module as well. In addition, since the monic polynomial displayed
above has coefficients in R[s0 , . . . , sn−1 ], t is integral over R[s0 , . . . , sn−1 ], so
R[s0 , . . . , sn−1 , t] is a finitely generated R-module. Hence, t is integral over R
by Proposition C.0.21. 

If either ring in an integral extension is a field, then the integral extension

is simply a field extension.

C.0.23 Proposition. Let R be a ring and S be an integral domain that is

integral over R. Then, R is a field if and only if S is a field.

22
Proof: Suppose R is a field and fix a non-zero s ∈ S. Then, s is integral over
R, so there exist a0 , a1 , . . . , an−1 ∈ R such that

sn + an−1 sn−1 + · · · + a1 s + a0 = 0.

Since S is an integral domain, we may assume a0 6= 0, as otherwise s factors

out of the left-hand side of the equation, implying that s is a zero divisor. Then

s(sn−1 + an−1 sn−2 + · · · + a1 ) = −a0 ,

and since (−1/a0 ) ∈ R, this shows that

s(sn−1 + an−1 sn−2 + · · · + a1 )(−1/a0 ) = −a0 (−1/a0 ) = 1,

so s is invertible. Therefore, S is a field.

Conversely, suppose that S is a field and fix a non-zero r ∈ R. Since r−1 ∈ S
is integral over R, there exist a0 , a1 , . . . , an−1 ∈ R such that

r−m + am−1 r−m+1 + · · · + a1 r−1 + a0 = 0.

Then
r−1 = −(am−1 + · · · + a1 rm−2 + a0 rm−1 ) ∈ R.
Therefore, R is a field. 

Most rings we deal with in these notes are also vector spaces over some
field k, where the vector space addition is the same as addition in the ring,
and the scalar multiplication coincides with multiplication in the ring, after
identifying the scalar α ∈ k with the element α · 1 of the ring. Such rings are
called k-algebras. We deal with k-algebras elsewhere in these notes, but repeat
some basic facts about them here.
C.0.24 Examples.
(i) k[x1 , . . . , xn ] is a k-algebra.
(ii) Let A be a k-algebra and I an ideal of A. Then I is also a vector subspace
of A, and the ring quotient of A by I agrees with the vector space quotient
of A by I. Hence A/I is also a k-algebra.

Given k-algebras A and B, we define a k-algebra homomorphism from A to

B to be a ring homomorphism ϕ : A → B such that ϕ(α) = α for every α ∈ k.
C.0.25 Examples.
(i) Let A be a k-algebra and I be an ideal of A. The quotient map ϕ : A →
A/I is then a k-algebra homomorphism.
(ii) The map ϕ : C[x1 , . . . , xn ] → C[x1 , . . . , xn ] defined by

ϕ(an xn + · · · + a1 x + a0 ) = an xn + · · · + a1 x + a0

is a ring homomorphism that is not a C-algebra homomorphism.

23
 Let A be a k-algebra. We say that A is finitely generated if A = k[a1 , . . . , an ]
for some a1 , . . . , an ∈ A. Moreover, y1 , . . . , yn ∈ A are said to be algebraically
independent if there is no non-zero f ∈ k[x1 , . . . , xn ] such that f (y1 , . . . , yn ) =
0, or equivalently, if the k-algebra homomorphism

ϕ : k[x1 , . . . , xn ] → k[y1 , . . . , yn ]

defined by ϕ(xi ) = yi is an isomorphism.

C.0.26 Theorem (Noether Normalization Lemma). Let A be an inte-

gral domain that is a finitely generated k-algebra, and let d be the transcen-
dence degree of the fraction field of A over k. There exist elements y1 , . . . , yd ∈
A which are algebraically independent over k and are such that A is integral
over k[y1 , . . . , yd ].

Proof: Suppose that A = k[r1 , . . . , rn ]. If r1 , . . . , rn are already algebraically

independent over k, then we are done. If not, there is a non-trivial polynomial
relation amongst the ri ’s, i.e. there exists a finite family F of distinct tuples
j = (j1 , . . . , jn ) of non-negative integers and corresponding non-zero cofficients
aj ∈ k such that
X
aj r1j1 . . . rnjn = 0.
j∈F

Let m = (1, m2 , . . . , mn ) be a vector of positive integers, and let

y2 = r2 − r1m2 , . . . , yn = rn − r1mn .

We use the dot product j · m to denote j1 + m2 j2 + · · · + mn jn . Substituting

ri = yi + r1mi into the above relation, we get
X
aj r1j·m + f (r1 , y2 , . . . , yn ) = 0,
j∈F

where f is a polynomial in which no pure power of r1 appears. Let l be an inte-

ger greater than any component of a vector in F, and let m = (1, l, l2 , . . . , ln−1 ).
Then, all j·m are distinct for those j such that aj 6= 0. In this way, we obtain an
integral equation for r1 over k[y2 , . . . , yn ], implying that k[y2 , . . . , yn ][r1 ] is in-
tegral over k[y2 , . . . , yn ] by Proposition C.0.21. However, A = k[r1 , y2 , . . . , yn ]
by the definition of the yi ’s; it follows that A is integral over k[y2 , . . . , yn ].
We now proceed inductively, applying the same construction to k[y2 , . . . , yn ]
and using the transitivity of integral extensions, to shrink the number of y’s
down to an algebraically independent set. Thus, there exist algebraically in-
dependent elements y1 , . . . , yk ∈ A such that A is integral over k[y1 , . . . , yk ].
Since y1 , . . . , yk are still algebraically independent in the fraction field of A,
and since A is integral over k[y1 , . . . , yk ], they generate the fraction field of A
over k. Therefore, y1 , . . . , yk form a transcendence basis of the fraction field of
A over k, and k = d as desired. 

24
 In the particular case of fields, the Noether Normalization Lemma can be
restated more naturally in the language of field extensions.

C.0.27 Corollary. Let L be a field extension of k that is finitely generated

as a k-algebra. Then L is algebraic over k.

Proof: By the Noether Normalization Lemma, there exist algebraically in-

dependent y1 , . . . , yn ∈ L such that L is integral over k[y1 , . . . , yn ]. Since L
is a field, by Proposition C.0.23, k[y1 , . . . , yn ] is also a field. But this implies
k[y1 , . . . , yn ] = k, as no polynomial ring over a field is a field. Therefore, L is
integral over k, i.e. L is algebraic over k. 

C.0.28 Theorem (Weak Nullstellensatz). Suppose k is algebraically closed.

Then every maximal ideal in k[x1 , . . . , xn ] is of the form hx1 − a1 , . . . , xn − an i
for some (a1 , . . . , an ) ∈ An . Moreover, if I is a proper ideal of k[x1 , . . . , xn ]
then V(I) 6= ∅.

Proof: Let R = k[x1 , . . . , xn ] and let I be a maximal ideal of R. Then R/I is

a field extension of k and finitely generated as a k-algebra. By Corollary C.0.27,
it is an algebraic extension of k. Since k is algebraically closed, R/I = k. Let
ϕ : R → k be the quotient map, and let ai = ϕ(xi ) for i = 1, . . . , n. Let
I 0 = hx1 − a1 , . . . , xn − an i. For i = 1, . . . , n, we have

ϕ(xi − ai ) = ϕ(xi ) − ϕ(ai ) = ai − ϕ(ai ) = ai − ai = 0,

so ϕ(f ) = 0 for every f ∈ I 0 . Hence I 0 ⊆ I. However, I 0 is maximal, so I = I 0 .

Let I be a proper ideal of k[x1 , . . . , xn ]. Then I is contained in a maximal
ideal, so there exist a1 , . . . , an ∈ k such that

I ⊆ hx1 − a1 , . . . , xn − an i.

Thus
{(a1 , . . . , an )} = V(hx1 − a1 , . . . , xn − an i) ⊆ V(I),
so (a1 , . . . , an ) ∈ V(I), and V(I) 6= ∅. 

C.0.29 Theorem (Strong Nullstellensatz). Suppose √ k is algebraically

√ closed,
and let I ⊆ k[x1 , . . . , xn ] be an ideal. Then I(V(I))
√ = I, or I = I, so I is
the ideal of a set of points if and only if I = I.
√
Proof: Since I ⊆ I(V(I)), we only need to prove the reverse inclusion. Fix
g ∈ I(V(I)). By the Hilbert Basis Theorem, we have I = hf1 , . . . , fm i for some
f1 , . . . , fm ∈ k[x1 , . . . , xn ]. Let us introduce a new variable xn+1 and consider
the ideal I 0 = hf1 , . . . , fm , 1−xn+1 gi of k[x1 , . . . , xn , xn+1 ]. If f1 , . . . , fm vanish
at (a1 , . . . , an+1 ) ∈ An+1 , then g also vanishes at (a1 , . . . , an+1 ), as g ∈ I(V(I)),
so 1 − xn+1 g is non-zero. Hence V(I 0 ) = ∅. By the Weak Nullstellensatz, I 0

25
is not proper, i.e. 1 ∈ I 0 . Therefore, there exist h1 , . . . , hn+1 ∈ k[x1 , . . . , xn+1 ]
such that
1 = h1 f1 + · · · + hn fn + hn+1 (1 − xn+1 g).
Working in the field of fractions of k[x1 , . . . , xn+1 ], substitute g −1 for xn+1 and
multiply both sides by an appropriate power g k of g to clear denominators on
the right-hand side and give

g k = h̃1 f1 + · · · + h̃n fn + h̃n+1 (1 − g −1 g) = h̃1 f1 + · · · + h̃n fn ∈ I,

where h̃i (x1√, . . . , xn ) := g k hi (x1 , . . . , x√n , g −1 ) ∈ k[x1 , . . . √

, xn ] for all i = 1, . . . , n.
Hence g ∈ I. Therefore, I(V(I)) ⊆ I and I(V(I)) = I. 

26