Chapter 13
Chapter 13
algebra.
1. {xi} ∈ ∆ for 1 ≤ i ≤ n
2. If F ∈ ∆ and G ⊆ F , then G ∈ ∆.
1
below concerning empty simplicial complexes and empty faces).
An i-dimensional face is called an i-face. Soon we will see the
geometric reason for our definition of dimension.
2
A set {v0, v1, . . . , vj } ⊂ Rd is affinely independent if the fol-
lowing condition holds: if α0 , α1 , . . . , αj are real numbers for
P P
which αi vi = 0 and αi = 0, then α0 = α1 = · · · = αj = 0.
Equivalently, define an affine subspace of Rd to be the translate
of a linear subspace, i.e., a set
A = {v ∈ Rd : v · y (1) = α1 , . . . , y (k) = αk },
3
A (finite) geometric simplicial complex is a finite set Γ of
simplices in Rd such that the following two conditions hold:
S
We sometimes identify Γ with the union σ∈Γ σ of its simplices.
In this situation Γ is just a subset of Rd , but it is understood
that it has been described as a union of certain simplices.
4
However, for our purposes we don’t need this uniqueness.
13.3 Example. Let ∆ = h123, 234, 235, 36, 56, 57, 8i. A
geometric realization of ∆ is shown in Figure 1, projected from
three dimensions. Note that since three triangles share the edge
23, any geometric realization in Rd requires d ≥ 3. It is a result
of Karl Menger, though irrelevant for us, that any d-dimensional
simplicial complex can be realized in R2d+1 , and that this result
is best possible, i.e., the dimension 2d + 1 cannot in general
be decreased. In fact, the simplicial complex whose facets are
all the (d + 1)-element subsets of a (2d + 3)-element set cannot
be realized in R2d . For example, when d = 1 we get that the
complete graph K5 cannot be embedded in the plane (without
crossing edges), a famous result in graph theory known to Euler
at least implicitly, since he showed in 1750 that f1 ≤ 3f0 − 6 for
any planar graph. The first person to realize explicitly that K5
is not planar seems to be A. F. Möbius in 1840, who stated the
result in the form of a puzzle.
5
4
7
5
2
8
1 3 6
Figure 1: A geometric realization
3 2
3
2
1
Figure 2: The boundary of an octahedron
6
We now come to the combinatorial information about simpli-
cial complexes that is our primary interest in this chapter. For
i ≥ −1, let fi be the number of i-dimensional faces of ∆. Thus
f−1 = 1 unless ∆ = Ø, and f0 = #V , the number of vertices of
∆. If dim ∆ = d − 1, then the vector
f (∆) = (f0, f1, . . . , fd−1)
is called the f -vector of ∆. Thus the simplicial complex ∆ of
Figure 1 has f -vector (8, 10, 3), while that of Figure 2 has f -
vector (6, 12, 8).
7
Proof. The proof is based on the following simple combinatorial
identity. Let 1 ≤ i ≤ m. Then
m m−1 m−i+1 m+1
+ + ··· + +1= . (2)
i i−1 1 i
This identity can easily be proved by induction on i, for instance.
It also has a simple combinatorial interpretation. Namely, the
right-hand side is the number of i-element subsets S of the set
[m+1] = {1, 2, . . . , m+1}. The number of such subsets
for which
m−s
the least missing element is s + 1 is equal to i−s . Summing
over all 0 ≤ s ≤ i completes the proof of equation (2).
We now prove
the proposition
by induction on j. For j = 1 we
n m
have n = 1 , while n 6= 1 for m 6= n. Hence the proposition
is true for j = 1.
9
provide some intuition and at the same time convey some idea
of the proof.
10
with a1 < · · · < aj . Then
aj aj−1 a1
n= + + ··· + ,
j j−1 1
and this formula gives the j-binomial expansion of n.
11
Now suppose that f = (f0, . . . , fd−1) ∈ Pd . Define a collection
Γf of subsets of N to consist of the empty set Ø together with
the first fi of the (i + 1)-element subsets of N in rlex order. For
example, if f = (6, 8, 5, 2) then (writing as usual {1, 2, 3} = 123,
etc.)
Proof. Let us use the notation [0, m] = {0, 1, . . . , m} and for any
set S,
S
= {T ⊆ S : #T = k}.
k
ni
Let fi = ni+1i+1
+ i +· · ·+ n11 be the (i+1)-binomial expansion
of fi . By the definition of rlex order, we see that the set X of
the first fi (i + 1)-elements of N in rlex order is given by
[0, ni+1 − 1] [ [0, ni − 1]
X= {ni+1} ∪
i+1 i
[
[0, ni−1 − 1]
[
{ni+1, ni} ∪ ··· .
i−1
The set of (i + 2)-elements subsets F of N all of whose (i + 1)-
element subsets belong to X is given by
[0, ni+1 − 1] [ [0, ni − 1]
X= {ni+1} ∪
i+2 i+1
12
[
[0, ni−1 − 1]
[
{ni+1, ni} ∪ ··· .
i
(i+1)
These are just the first fi (i + 2)-element subsets of N in rlex
order, and the proof follows.
Γf = {Ø, 0, 1, 2, 3, 4, 01, 02, 12, 03, 13, 23, 04, 012, 013, 023, 123, 014}.
13
but 146∈ Γf . Hence (5, 7, 5) is not an f -vector. In fact, we have
4 1 (2) 4 1
7 = 2 + 1 and 7 = 3 + 2 = 4 < 5.
14
Note. Let ∆ be a pure (d − 1)-dimensional simplicial com-
plex. It is easy to see (Exercise 2) that a facet ordering F1, . . . Ft
is a shelling if and only if for all 2 ≤ i ≤ t, the subcomplex
hF1 , . . . , Fi−1i ∩ hFi i (i.e., the set of faces of Fi that already be-
long to hF1 , . . . , Fi−1i) is a pure simplicial complex of dimension
d − 2. More informally, Fi attaches along some nonempty union
of its facets.
13.11 Example.
15
1
a b
1 2 3
4 2
a b c d
d c
3
(a) (b)
a c a c
e
1 2 1 2
b d b d
(c) (d)
16
d
8
2
5
a b
7
1
4 3
c
6
f e
17
Figure 5: A nonshellable simplicial complex
18
and h-vector contain equivalent information—f (∆) determines
h(∆) and vice versa.
13.14 Example.
19
There are some elementary properties of the h-vector worth
noting:
20
say that χ̃(Γ) is a topological invariant of Γ since it depends
only on the geometric realization |Γ| as a topological space.
Equation (6) therefore shows that
hd = (−1)d−1χ̃(Γ). (7)
χ̃(X) = χ(X) − 1
21
P
contribution to the polynomial di=0
fi−1(x −1)d−i from adjoin-
ing Fj is (using the symmetry d−m
i
d−m
= d−m−i and the binomial
theorem) is given by
X
d−m
d−m
X d − m
d−m
(x − 1)d−(m+i) = (x − 1)i
i=0
i i=0
i
= xd−m ,
22
The condition of Corollary 13.16 is necessary but not suf-
ficient for shellability. For instance, the disjoint union of two
cycles (a one-dimensional simplicial complex) satisfies hd = 1
but isn’t shellable. (See Exercise 24.) For some more subtle
examples, see Exercises 9 and 10.
23
generated by the minimal nonfaces of ∆, that is, those nonfaces
for which no proper subset is a nonface. A minimal nonface is
also called a missing face.
For Figure 1, the missing faces all have two elements except
for {3, 5, 6}. For the octahedron of Figure 2, the missing faces
are (writing as usual 11′ for {1, 1′}, etc.) 11′, 22′, and 33′.
The quotient ring K[∆] := K[V ]/I∆ is called the face ring
(also called the Stanley-Reisner ring) of ∆. It is the fundamental
algebraic object of this chapter.
24
we identify elements of K[V ] with their images in K[∆].) For
i ≥ 0 define K[∆]i to be the span of all monomials u of degree
i satisfying supp(u) ∈ ∆. Then
λ#F
= . (10)
(1 − λ)#F
In particular, when F = Ø the two sides of equation (10) are
25
equal to 1. Summing over all F ∈ ∆ gives
X λ#F
L(K[∆], λ) =
(1 − λ)#F
F ∈∆
X d
λi
= fi−1
i=0
(1 − λ)i
Xd
fi−1λi (1 − λ)d−i
i=0
= .
(1 − λ)d
Now
d
X d
X d−i
i d−i d 1
fi−1λ (1 − λ) = λ fi−1 −1
i=0 i=0
λ
Xd
= λd hi λ−(d−i) (by (4))
i=0
d
X
= hi λi ,
i=0
26
are algebraically independent over K, and is also the length ℓ of
the longest chain p0 ⊂ p1 ⊂ · · · ⊂ pℓ of prime ideals of K[∆].
27
Proof. If θ is an NZD then by equation (11) we have
H(K[∆]/(θ), i + 1) = H(K[∆], i + 1) − H(K[∆], i).
Multiplying both sides by λi+1 and summing on i ≥ −1 gives
L(K[∆]/(θ), λ) = L(K[∆], λ) − λL(K[∆], λ),
so
L(K[∆]/(θ), λ)
L(K[∆], λ) = .
1−λ
If θ is not an NZD, then we always have
dimK θK[∆]i ≤ dimK K[∆]i,
and for at least one i strict inequality holds. This is easily seen
to imply that strict inequality holds in equation (13).
28
It is easy to see that a regular sequence θ1, . . . , θj ∈ K[∆]1
is algebraically independent over K (Exercise 19). In other
words, there does not exist a polynomial 0 6= P (t1 , . . . , tk ) ∈
K[t1, . . . , tk ] for which P (θ1, . . . , θk ) = 0 in K[∆]. Let us point
out that if the sequence θ1, . . . , θj ∈ K[∆] is algebraically inde-
pendent and moreover each θi is an NZD in K[∆], then these
conditions are not sufficient for θ1 , . . . , θj to be a regular se-
quence. For instance, if ∆ has vertices a, b, c and the single edge
ab, then a−c and b−c are algebraically independent NZDs. How-
ever, in the ring K[∆]/(a − c) we have c 6= 0 but (b − c)c = 0. In
fact, we have depth K[∆] = 1, e.g., by Exercise 22. For another
example, let ∆ have vertices a, b, c and edges ab, bc, and assume
that char(K) 6= 2. Now a + b and a − b are algebraically inde-
pendent NZDs but not a regular sequence since in K[∆]/(a + b)
we have c 6= 0 and c(a − b) = 0. Unlike the previous example,
this time we have depth K[∆] = 2. A regular sequence of length
two is given by, for instance, a − c, b.
L(R, λ) = h0 + h1 λ + · · · + hd λd , (14)
29
13.21 Definition. Assume that K is an infinite field.
We say that the simplicial complex ∆ is Cohen-Macaulay (with
respect to the field K) and that the ring K[∆] is a Cohen-
Macaulay ring if dim K[∆] = depth K[∆].
30
Let us first consider the second question. A multicomplex
Γ on a set V is a multiset analogue of a simplicial complex
whose vertex set is contained in V . More precisely, Γ is a col-
lection of multisets (sets with repeated elements, as discussed
on page 1 of Algebraic Combinatorics), such that every ele-
ment of Γ is contained in V , and if M ∈ Γ and N ⊆ M,
then N ∈ Γ. We will assume from now on that the underly-
ing set V is finite. For example (writing 112 for {1, 1, 2}, etc.),
Γ = {Ø, 1, 2, 3, 11, 12, 112, 1112} is not a multicomplex, since
1112 ∈ Γ and 111 ⊆ 1112, but 111 6∈ Γ.
31
(e.g., Theorem 13.25). Moreover, e-vectors of multicomplexes
coincide with the sequence of Hilbert function values of stan-
dard graded K-algebras (not defined here, though K[∆] and its
quotients considered here are special cases), so one can think
that h stands for “Hilbert.” To avoid any possible confusion we
will use the new notation ei and terminology e-vector.
While we need the next result only for quotients of face rings
by a regular sequence, it involves no extra work to prove it in
much greater generality. For this purpose, we define a homoge-
neous ideal of the polynomial ring K[x1, . . . , xn] to be an ideal
I generated by homogeneous polynomials.
32
follows: define
R
xa11 · · · xann < xb11 · · · xbnn ,
P P
where ai = bi, if
R
(1a1 , . . . , nan ) < (1b1 , . . . , nbn ),
33
R
monomials w < u, say
X
u= αw w.
R
w<u
R
Multiply both sides by v/u. It is easy to see that if w < u then
R
wv/u < v. Thus we have expressed v as a linear combination of
R
monomials wv/u < v, contradicting v ∈ Bj .
R = R0 ⊕ R1 ⊕ · · · = K[∆]/(θ1, . . . , θd).
34
Cohen-Macaulay for any infinite field K. Moreover, if F1 , . . . , Ft
is a shelling of ∆ with restrictions G1 , . . . , Gt , then xG1 , . . . , xGt
is a K-basis for R = K[∆]/(θ1, . . . , θd ) for any regular sequence
θ1, . . . , θd ∈ K[∆]1.
ψi = θi |xj =0 if xj 6∈F ,
then ψ1 , . . . , ψd span KF .
35
a one-term sequence) is a shelling of ∆ with G1 = Ø. Since
xØ = 1 is a basis for K, the theorem is true for t = 1.
36
Now let R′ = R/(xGt ) and ∆t−1 = hF1 , . . . , Ft−1i. By defini-
tion of Gt we have
K[∆t−1] = K[∆]/(xGt ).
Condition (P) still holds for K[∆t−1] (since the facets of ∆t−1
are also facets of ∆). Moreover,
R′ = K[∆t−1]/(θ1, . . . , θd ).
37
(a) There exists a (d − 1)-dimensional Cohen-Macaulay simpli-
cial complex (over any infinite field) ∆ with h(∆) = h.
(b) There exists a (d − 1)-dimensional shellable simplicial com-
plex ∆ with h(∆) = h.
(c) The sequence h is an e-vector.
38
We leave as an exercise (Exercise 25) to show that σ is a
shelling of a (d−1)-dimensional simplicial complex ∆ on the ver-
tex set {1, 2, . . . , h1 +d}. Moreover, if we write σ = (F1, . . . , Fm)
P
(where m = hi = fd−1(∆)) and if Fk is given by the sequence
on the right-hand side of equation (16), then the restriction Gk
is given by Gk = {a1 + d − i + 1, a2 + d − i + 2, . . . , ai + d}.
This being the case, exactly hi restrictions Gk have i elements,
so indeed h(∆) = h.
39
e i (lk∆ (F ); K) =
• For every F ∈ ∆ (including F = Ø), we have H
0 for all i 6= dim lk∆ (F ).
40
bottom and top. We now have the following exact analogue of
the Kruskal-Katona theorem.
41
The proof of the “only if” direction is similar to that of the
Kruskal-Katona theorem. A multicomplex as in Theorem 13.29
is called compressed. Given any multicomplex Γ, we transform it
by a sequence of simple operations into a compressed multicom-
plex, at all steps preserving the e-vector. We omit the details,
which are somewhat more complicated than in the simplicial
complex case.
42
by M.-P. Schützenberger in a rather obscure journal [15]. The
first published proofs were by J. Kruskal [9] and later indepen-
dently by G. O. H. Katona [7]. A nice survey of this area is given
by Greene and Kleitman, [6], including a good presentation of
a proof of the Kruskal-Katona theorem due to Clements and
Lindström [3].
43
Cohen-Macaulay rings are named after I. Cohen [4] and F. S.
Macaulay [10], who were interested in them primarily because of
their connection with “unmixedness” theorems. For a modern
treatment see the text of W. Bruns and J. Herzog [2].
References
44
[8] B. Kind and P. Kleinschmidt, Schälbare Cohen-Macaulay-
Komplexe und ihre Parametrisierung, Math. Z. 167 (1979),
173–179.
[9] J. Kruskal, The number of simplices in a a complex, Math-
ematical Optimization Techniques, University of California
Press, Berkeley and Los Angeles, 1963, 251–278.
[10] F. S. Macaulay, The Algebraic Theory of Modular Systems,
Cambridge University Press, Cambridge, 1916; reprinted by
Cambridge University Press, Cambridge/New York, 1994.
[11] F. S. Macaulay, Some properties of enumeration in the
theory of modular systems, Proc. London Math. Soc. 26
(1927), 531–555.
[12] J. Matouss̆ek, Using the Borsuk-Ulam theorem: Lectures
on Topological Methods in Combinatorics and Geometry,
Universitext, Springer-Verlag, Berlin, 2003.
[13] K. Menger, Untersuchungen über allgeneine Metrik?, Math.
Ann. 100 (1928), 75–163.
[14] G. Reisner, Cohen-Macaulay quotients of polynomial rings,
Advances in Math. 21 (1976), 30–49.
[15] M.-P. Schützenberger, A characteristic property of certain
polynomials of E. F. Moore and C. E. Shannon, in RLE
Quarterly Progress Report, No. 55, Research Laboratory
of Electronics, M.I.T., 1959, 117–118.
[16] E. Sperner, Über einen kombinatorischen Satz von
Macaulay und seine Anwendung auf die Theorie der Poly-
nomideale, Abh. Math. Sem. Univ. Hamburg 7 (1930), 149–
163.
45
[17] R. Stanley, Cohen-Macaulay rings and constructible poly-
topes, Bull. Amer. Math. Soc. 81 (1975), 133–135.
[18] R. Stanley, The Upper Bound Conjecture and Cohen-
Macaulay rings, Studies in Applied Math. 54 (1975), 135–
142.
[19] R. Stanley, Combinatorics and Commutative Algebra,
second ed., Progress in Mathematics 41, Birkhäuser,
Boston/Basel/Berlin, 1996.
[20] F. Whipple, On a theorem due to F. S. Macaulay, J. London
Math. Soc. 28 (1928), 431–437.
46
Exercises for Chapter 13
47
7. Prove the assertion of Example 13.11(d) that a simplicial
complex ∆ is shellable if and only if the cone C(∆) is
shellable.
8. A matroid complex is a simplicial complex ∆ on the vertex
set V such that for any W ⊆ V , the restriction ∆W of ∆
to W , i.e.,
∆W = {F ∈ ∆ : F ⊆ W },
is pure.
(a) Let V = {x1, . . . , xn} be a set of distinct nonzero vec-
tors in some vector space over a field. Define
∆ = {F ⊆ V : F is linearly independent}.
48
(c) (even more difficult) Find a triangulation of a three-
dimensional sphere that is not shellable.
11. (*) A partial shelling of a pure (d − 1)-dimensional complex
is a sequence F1 , . . . , Fr of some subset of the facets such
that this sequence is a shelling order for the simplicial com-
plex hF1 , . . . , Fr i which they generate. Clearly if F1 , . . . , Ft
is a shelling of ∆, then F1, . . . , Fr is a partial shelling for all
1 ≤ r ≤ t. Give an example of a shellable simplicial com-
plex that has a partial shelling that cannot be extended to
a shelling.
12. Let (f0, f1, . . . , fd−1) be the f -vector of a (d−1)-dimensional
simplicial complex ∆. We will illustrate a certain procedure
with the example (6, 12, 8) (the f -vector of an octahedron).
Write down the numbers f0, f1, . . . , fd−1 on a diagonal, and
put 1 to the left of f0:
1 6
12
8
Think of the 1 as being preceded by a string of 0’s. Turn
the array into a difference table by writing below each pair
49
of consecutive numbers their difference:
1 6
1 5 12
1 4 7 8
Now write down one further row of differences:
1 6
1 5 12
1 4 7 8
1 3 3 1
Show that this bottom row is the h-vector of ∆.
13. Find the f -vector and h-vector of the simplicial complex
whose geometric realization is the boundary of an icosahe-
dron.
14. Let ∆d be the simplicial complex on the vertex set V =
{x1, . . . , xd , y1 , . . . , yd} whose faces are those subsets of V
that do not contain both xi and yi , for any 1 ≤ i ≤ d.
Compute the h-vector of ∆d .
Cultural note. Let δi be the ith unit coordinate vector
in Rd . A geometric realization of ∆d consists of the bound-
ary of the convex hull Cd of the vectors ±δi , 1 ≤ i ≤ d. The
polytope C is called the d-dimensional cross-polytope and is
a d-dimensional generalization of an octahedron, the case
d = 3.
15. Give an example of two simplicial complexes ∆1 and ∆2
such that the geometric realizations of ∆1 and ∆2 are home-
omorphic, the h-vector of ∆1 is nonnegative, and some
hi (∆2) < 0. What is the smallest possible dimension of
∆1 and ∆2?
50
16. (difficult from first principles) (*) Let ∆ be a triangulation
of a (d−1)-dimensional sphere, and let h(∆) = (h0 , h1 , . . . , hd ).
Show that hi = hd−i for 0 ≤ i ≤ d. This result is called the
Dehn-Sommerville equations for spheres.
17. Let h = (h0 , h1, . . . , hd ) and k = (k0, k1, . . . , kd ) be e-
vectors. Define
h ∧ k = (min{h0 , k0}, min{h1 , k1}, . . . , min{hd , kd })
h ∨ k = (max{h0 , k0}, max{h1 , k1}, . . . , max{hd , kd }).
Show that h ∧ k and h ∨ k are e-vectors.
18. (*) Suppose that Γ and ∆ are simplicial complexes whose
face rings K[Γ] and K[∆] are isomorphic as K-algebras (or
even as rings). Show that Γ and ∆ are isomorphic.
19. Show that a regular sequence θ1, . . . , θj ∈ K[∆]1 is alge-
braically independent over K.
20. (a) Let θ1 , . . . , θj ∈ K[∆]1 be a regular sequence. Show
that any permutation of this sequence is also a regular
sequence.
(b) Show that each θi is an NZD in K[∆].
21. Let ∆ be any (d − 1)-dimensional simplicial complex, and
let θ1, . . . , θd ∈ K[∆]1. Show that the quotient ring R =
K[∆]/(θ1, . . . , θd ) is a finite-dimensional vector space over
K if and only if θ1, . . . , θd satisfy Property (P).
22. Show that the face ring K[∆] of a simplicial complex ∆ has
depth one if and only if ∆ is disconnected. Deduce that a
disconnected simplicial complex of dimension at least one
is not Cohen-Macaulay.
51
23. Let Γ and ∆ be simplicial complexes on disjoint vertex sets
V and W , respectively. Define the join Γ ∗ ∆ to be the
simplicial complex on the vertex set V ∪W with faces F ∪G,
where F ∈ Γ and G ∈ ∆. (If Γ consists of a single point,
then Γ ∗ ∆ is the cone over ∆. If Γ consists of two disjoint
points, then Γ ∗ ∆ is the suspension of ∆.)
(a) Compute the h-vector h(Γ ∗ ∆) in terms of h(Γ) and
h(∆).
(b) Show that if Γ and ∆ are Cohen-Macaulay, then so is
Γ ∗ ∆.
(c) Generalizing Exercise 7, show that if Γ and ∆ are shellable,
then so is Γ ∗ ∆.
24. Let ∆ be a one-dimensional simplicial complex. Show that
the following three conditions are equivalent: (a) ∆ is Cohen-
Macaulay, (b) ∆ is shellable, and (c) ∆ is connected.
25. Complete the proof of Theorem 13.25 by showing that the
sequence σ is a shelling of ∆ with the stated restrictions
Gk .
26. (*) Let ∆ be a four-dimensional shellable simplicial complex
with f0 = 13, f1 = 50, and f2 = 129. What is the most
number of facets that ∆ can have?
27. (*) Let ∆ be a (d − 1)-dimensional Cohen-Macaulay sim-
plicial complex with h-vector (h0 , h1, . . . , hd ). Let ∆′ be
a (d − 1)-dimensional Cohen-Macaulay subcomplex of ∆
with h-vector (h′0 , h′1, . . . , h′d ). Show that h′i ≤ hi for all
0 ≤ i ≤ d.
52
28. (*) Let ∆ be a (d − 1)-dimensional simplicial complex on
the vertex set V . We say that ∆ is balanced if we can write
V as a disjoint union V = V1 ∪ V2 ∪ · · · ∪ Vd such that for
every F ∈ ∆ and every 1 ≤ i ≤ d we have #(F ∩ Vi ) ≤ 1.
In particular, if ∆ is pure then always #(F ∩ Vi ) = 1.
(Sometimes ∆ is required to be pure in the definition of
balanced.) Suppose that (h0, h1 , . . . , hd ) is the h-vector of
a Cohen-Macaulay balanced simplicial complex ∆. Show
that (h1 , h2, . . . , hd ) is the f -vector of a balanced simplicial
complex. You may assume the following result: let ∆ be a
Cohen-Macaulay simplicial complex of dimension d − 1. If
θ1, . . . , θd ∈ K[∆]1 satisfies (P) if and only if θ1, . . . , θd is a
regular sequence.
53
13.26. Answer: 325.
13.27. Let I be the ideal of K[∆] generated by all monomials xF ,
where F 6∈ ∆′. Clearly K[∆′] is isomorphic to K[∆]/I (as
a K-algebra). Let θ1, . . . , θd ∈ K[∆]1 satisfy Property (P),
and consider the natural map
54
Index
affine span, 3 reduced, 20
affine subspace, 3 Euler, Leonhard, 5
dimension, 3 e-vector (of a multicomplex), 30
affinely independent, 3
face
balanced, see simplicial complex, bal- missing, 23
anced of a simplex, 3
Bruggesser, Heinz, 42 of a simplicial complex, 1
Bruns, Winfried, 43 face ring, see ring, face
facet, 1
Clements, George F., 42, 43 Flores, A., 42
Cohen, Irvin Sol, 43 f -vector, 7
Cohen-Macaulay face ring, see ring,
face, Cohen-Macaulay geometric realization, 4
Cohen-Macaulay simplicial complex, Greene, Curtis, 42
see simplicial complex, Cohen-
Macaulay Herzog, Jürgen, 43
compressed multicomplex, see multi- Hilbert series, 24
complex, compressed Hochster, Melvin, 42
compressed simplicial complex, see sim-homogeneous ideal, 32
plicial complex, compressed h-vector (of a simplicial complex), 18
cone, 16, 51 i-face, 2
convex hull, 2
convex set, 2 join (of simplicial complexes), 51
cross-polytope, 50
Katona, Gyula O. H., 9, 42
Dehn-Sommerville equations, 50 Kind, Bernd, 43
depth, 28 Kleinschmidt, Peter, 43
dimension Kleitman, Daniel J., 42
Krull, 26 Krull dimension, see dimension, Krull
of a face, 1 Kruskal, Joseph Bernard, 9, 42
divides (as a relation on monomials), Kruskal-Katona theorem, 9
31
dunce hat, topological, 47 Lindström, Bernt, 42, 43
link (of a face of a simplicial com-
Euler characteristic, 20 plex), 38
55
Macaulay, Francis Sowerby, 30, 43 Schützenberger, Marcel-Paul, 9, 42
Mani, Peter, 42 shellable, see simplicial complex, shellable
Matous̆ek, Jir̄ı́, 42 shelling, 14
matroid complex, 46 partial, 48
Menger, Karl, 5, 42 shelling order, 14
minimal nonface, see nonface, mini- simplex, 3
mal dimension, 3
missing face, see face, missing simplices, 3
Möbius, August Ferdinand, 5 simplicial complex, 1
multicomplex, 30 abstract, 1
compressed, 41 balanced, 52
M-vector, 30 Cohen-Macaulay, 29
compressed, 13
nonface, 23 geometric, 4
minimal, 23 pure, 14
non-zero-divisor, 26 shellable, 14
NZD, 26 Sperner, Emanuel, 43
octahedron, 5 Stanley, Richard Peter, 42
order ideal of monomials, 30, 31 Stanley-Reisner ring, see ring, Stanley-
O-sequence, 30 Reisner
subcomplex (of a simplicial complex),
partial shelling, see shelling, partial 14
pure, see simplicial complex, pure support (of a monomial), 24
suspension, 51
reduced Euler characteristic, see Eu-
ler characteristic, reduced topological dunce hat, see dunce hat,
regular sequence, 28 topological
Reisner, Gerald Allen, 38, 42 topological invariant, 20
restriction triangulation, 5
of a facet in a shelling, 14
of a simplicial complex, 46 unmixedness, 43
reverse lex order, 10 van Kampen, Egbert Rudolf, 42
on monomials, 32 vertex (of a geometric simplicial com-
reverse lexicographic order, 10 plex), 3
ring vertex set (of a simplicial complex), 1
face, 24
Cohen-Macaulay, 29 Whipple, Francis John Welsh, 43
Stanley-Reisner, 24
56