Math 2.0

Combinatorics Proof

Proof of Sperner's theorem by Lubell's chain counting

Builds on Binomial coefficientFactorial formula for binomial coefficientsNumber of k-element subsetsNumber of permutationsDouble counting

Proof

We count pairs of a full chain and a member of ๐’ฎ\mathcal{S} on it (Lubell's method). Then we examine when equality holds.

Setup.

  1. Let EE be a set with nn elements and let m=โŒŠn/2โŒ‹m = \floor{n/2}. Let ๐’ฎ\mathcal{S} be a family of subsets of EE in which no set is a proper subset of another. Write (nk)\binom{n}{k} for the binomial coefficient.
  2. A listing of EE is a sequence (x1,โ€ฆ,xn)(x_1, \dots, x_n) in which each element of EE appears exactly once. There are n!n! listings, by the number of permutations.
  3. A listing ฯƒ=(x1,โ€ฆ,xn)\sigma = (x_1, \dots, x_n) has the chain Cฯƒ={{x1,โ€ฆ,xk}:0โ‰คkโ‰คn}C_\sigma = \set{\set{x_1, \dots, x_k} : 0 \le k \le n}. It has exactly one set of each size kk, namely its first kk elements.

The binomial coefficients.

  1. By the factorial formula, (nk+1)/(nk)=(nโˆ’k)/(k+1)\binom{n}{k+1} \big/ \binom{n}{k} = (n-k)/(k+1) for 0โ‰คk<n0 \le k < n. This ratio is greater than 11 when 2k<nโˆ’12k < n - 1, equal to 11 when 2k=nโˆ’12k = n - 1, and less than 11 when 2k>nโˆ’12k > n - 1.
  2. So (nk)\binom{n}{k} strictly increases for 0โ‰คkโ‰คโŒŠn/2โŒ‹0 \le k \le \floor{n/2}, strictly decreases for โŒˆn/2โŒ‰โ‰คkโ‰คn\ceil{n/2} \le k \le n, and (nโŒŠn/2โŒ‹)=(nโŒˆn/2โŒ‰)\binom{n}{\floor{n/2}} = \binom{n}{\ceil{n/2}}.
  3. Hence (nk)โ‰ค(nm)\binom{n}{k} \le \binom{n}{m} for all 0โ‰คkโ‰คn0 \le k \le n, with equality exactly when kโˆˆ{โŒŠn/2โŒ‹,โŒˆn/2โŒ‰}k \in \set{\floor{n/2}, \ceil{n/2}}.

The LYM inequality.

  1. Let AโІEA \subseteq E with โˆฃAโˆฃ=k\abs{A} = k. Then AโˆˆCฯƒA \in C_\sigma exactly when the first kk entries of ฯƒ\sigma list AA and the others list Eโˆ–AE \setminus A. By Step 2, there are k!โ€‰(nโˆ’k)!k!\,(n-k)! such listings.
  2. Each chain CฯƒC_\sigma contains at most one member of ๐’ฎ\mathcal{S}. Indeed, of two distinct sets of CฯƒC_\sigma, one is a proper subset of the other.
  3. Double count the pairs (ฯƒ,A)(\sigma, A) with Aโˆˆ๐’ฎโˆฉCฯƒA \in \mathcal{S} \cap C_\sigma. By Step 7 there are โˆ‘Aโˆˆ๐’ฎโˆฃAโˆฃ!โ€‰(nโˆ’โˆฃAโˆฃ)!\sum_{A \in \mathcal{S}} \abs{A}!\,(n - \abs{A})! pairs. By Step 8 there are at most n!n! of them.
  4. Divide by n!n! and use the factorial formula: โˆ‘Aโˆˆ๐’ฎ1/(nโˆฃAโˆฃ)โ‰ค1\sum_{A \in \mathcal{S}} 1 \big/ \binom{n}{\abs{A}} \le 1.

The bound.

  1. By Step 6, each term satisfies 1/(nโˆฃAโˆฃ)โ‰ฅ1/(nm)1 \big/ \binom{n}{\abs{A}} \ge 1 \big/ \binom{n}{m}. So โˆฃ๐’ฎโˆฃ/(nm)โ‰คโˆ‘Aโˆˆ๐’ฎ1/(nโˆฃAโˆฃ)โ‰ค1\abs{\mathcal{S}} \big/ \binom{n}{m} \le \sum_{A \in \mathcal{S}} 1 \big/ \binom{n}{\abs{A}} \le 1.
  2. Hence โˆฃ๐’ฎโˆฃโ‰ค(nm)\abs{\mathcal{S}} \le \binom{n}{m}.

The two extremal families attain the bound.

  1. Let kโˆˆ{โŒŠn/2โŒ‹,โŒˆn/2โŒ‰}k \in \set{\floor{n/2}, \ceil{n/2}}, and let โ„’k\mathcal{L}_k be the family of all subsets of EE of size kk. Two distinct sets of the same size are not proper subsets of each other. So โ„’k\mathcal{L}_k has the property of ๐’ฎ\mathcal{S}.
  2. By the number of kk-element subsets and Step 5, โˆฃโ„’kโˆฃ=(nk)=(nm)\abs{\mathcal{L}_k} = \binom{n}{k} = \binom{n}{m}.

Equality forces an extremal family.

  1. Suppose now โˆฃ๐’ฎโˆฃ=(nm)\abs{\mathcal{S}} = \binom{n}{m}. Then both inequalities of Step 11 are equalities.
  2. The first equality and Step 6 show that every Aโˆˆ๐’ฎA \in \mathcal{S} has โˆฃAโˆฃโˆˆ{โŒŠn/2โŒ‹,โŒˆn/2โŒ‰}\abs{A} \in \set{\floor{n/2}, \ceil{n/2}}.
  3. The second equality means that the count of Step 9 is exactly n!n!. By Step 8, every chain CฯƒC_\sigma contains exactly one member of ๐’ฎ\mathcal{S}.
  4. If nn is even, every member of ๐’ฎ\mathcal{S} has size n/2n/2. So ๐’ฎโІโ„’n/2\mathcal{S} \subseteq \mathcal{L}_{n/2}, and both have (nm)\binom{n}{m} elements by Step 14. Hence ๐’ฎ=โ„’n/2\mathcal{S} = \mathcal{L}_{n/2}.
  5. Let nn be odd. Since (nm)โ‰ฅ1\binom{n}{m} \ge 1, the family ๐’ฎ\mathcal{S} has a member AA. Let k=โˆฃAโˆฃk = \abs{A}, so kโˆˆ{m,m+1}k \in \set{m, m+1}.
  6. Swap step. Let Aโˆˆ๐’ฎA \in \mathcal{S} with โˆฃAโˆฃ=k\abs{A} = k. Let Aโ€ฒ=(Aโˆ–{x})โˆช{y}A' = (A \setminus \set{x}) \cup \set{y} with xโˆˆAx \in A and yโˆ‰Ay \notin A, and put D=Aโˆ–{x}D = A \setminus \set{x}. We show Aโ€ฒโˆˆ๐’ฎA' \in \mathcal{S}.
  7. If k=mk = m, choose a listing that starts with the elements of DD, then yy, then xx. Its chain contains Aโ€ฒ=Dโˆช{y}A' = D \cup \set{y}, of size mm, and B=Dโˆช{x,y}B = D \cup \set{x, y}, of size m+1m + 1.
  8. By Steps 16 and 17, exactly one of Aโ€ฒA' and BB lies in ๐’ฎ\mathcal{S}. But AโŠŠBA \subsetneq B and Aโˆˆ๐’ฎA \in \mathcal{S}, so Bโˆ‰๐’ฎB \notin \mathcal{S}. Hence Aโ€ฒโˆˆ๐’ฎA' \in \mathcal{S}.
  9. If k=m+1k = m + 1, choose a listing that starts with the elements of DD, then yy. Its chain contains DD, of size mm, and Aโ€ฒ=Dโˆช{y}A' = D \cup \set{y}, of size m+1m + 1.
  10. Exactly one of them lies in ๐’ฎ\mathcal{S}. But DโŠŠAD \subsetneq A and Aโˆˆ๐’ฎA \in \mathcal{S}, so Dโˆ‰๐’ฎD \notin \mathcal{S}. Hence Aโ€ฒโˆˆ๐’ฎA' \in \mathcal{S}.
  11. Let Aโ€ฒโ€ฒA'' be any subset of EE of size kk, and let t=โˆฃAโˆ–Aโ€ฒโ€ฒโˆฃt = \abs{A \setminus A''}. Replacing elements one at a time, we get sets A=A0,A1,โ€ฆ,At=Aโ€ฒโ€ฒA = A_0, A_1, \dots, A_t = A'' of size kk, each obtained from the previous one as in Step 20.
  12. By Step 20 and induction on tt, Aโ€ฒโ€ฒโˆˆ๐’ฎA'' \in \mathcal{S}. So โ„’kโІ๐’ฎ\mathcal{L}_k \subseteq \mathcal{S}.
  13. By Step 14, โˆฃโ„’kโˆฃ=(nm)=โˆฃ๐’ฎโˆฃ\abs{\mathcal{L}_k} = \binom{n}{m} = \abs{\mathcal{S}}. Hence ๐’ฎ=โ„’k\mathcal{S} = \mathcal{L}_k, the family of all subsets of size โŒŠn/2โŒ‹\floor{n/2} or of all subsets of size โŒˆn/2โŒ‰\ceil{n/2}.
  14. By Steps 13, 14 and 15 to 27, equality holds exactly for these families.

Idea

A random ordering of EE builds a chain of n+1n+1 nested sets, and an antichain can meet it at most once. A set of size kk lies on a fraction 1/(nk)1/\binom{n}{k} of these chains, so these fractions add up to at most 11. For equality, every chain must meet the family at a middle size, and a swapping argument then fills a whole middle level.

Reading path

To understand this page from the ground up, read these 56 pages in order. The foundations come first. The path takes the proofs Proof of the number of k-element subsets by induction, Proof that a composition of injections is injective, Proof that a function is bijective iff it has a two-sided inverse, Proof of the complement principle, Proof of the number of injections, Proof of Pascal's rule from the recursive definition, Proof that a composition of surjections is surjective, Proof of the factorial formula for binomial coefficients by induction, Proof of the rule of sum, Proof of the number of permutations by counting injections, Proof of the bijection principle.

  1. Binomial coefficient definition
  2. Cardinality of a finite set definition
  3. Combination definition
  4. Double counting concept
  5. Element definition
  6. Factorial definition
  7. Falling factorial definition
  8. Permutation definition
  9. Proposition concept
  10. Proof of Pascal's rule from the recursive definition proof
  11. Set definition
  12. Conjunction definition
  13. Implication definition
  14. Predicate definition
  15. Pascal's rule statement
  16. Empty set object
  17. Ordered pair definition
  18. Intersection definition
  19. Existential quantifier definition
  20. Universal quantifier definition
  21. Proof of the factorial formula for binomial coefficients by induction proof
  22. Cartesian product definition
  23. Disjoint sets definition
  24. Subset definition
  25. Uniqueness quantifier definition
  26. Factorial formula for binomial coefficients statement
  27. Binary relation definition
  28. Function definition
  29. Codomain definition
  30. Function composition definition
  31. Identity function definition
  32. Injective function definition
  33. Restriction of a function definition
  34. Surjective function definition
  35. Left inverse definition
  36. Right inverse definition
  37. Proof that a composition of injections is injective proof
  38. Proof that a composition of surjections is surjective proof
  39. Bijective function definition
  40. Composition of injections is injective statement
  41. Composition of surjections is surjective statement
  42. Proof that a function is bijective iff it has a two-sided inverse proof
  43. Proof of the rule of sum proof
  44. Bijective iff it has a two-sided inverse statement
  45. Rule of sum statement
  46. Proof of the bijection principle proof
  47. Proof of the complement principle proof
  48. Bijection principle statement
  49. Complement principle statement
  50. Proof of the number of k-element subsets by induction proof
  51. Proof of the number of injections proof
  52. Number of k-element subsets statement
  53. Number of injections statement
  54. Proof of the number of permutations by counting injections proof
  55. Number of permutations statement
  56. Proof of Sperner's theorem by Lubell's chain counting proof

Where this leads

1 page builds on this one.

Further reading

  1. Web https://kam.mff.cuni.cz/~chvatal/est.pdf
  2. Web https://en.wikipedia.org/wiki/Sperner's_theorem, Proof
About this page
Tag57DD99JPR1
URIm2://57DD99JPR1
Kindproof
WritingW0 stub
MSC05D05
ProvesSperner's theorem
UsesBinomial coefficient, Factorial formula for binomial coefficients, Number of k-element subsets, Number of permutations, Double counting
Changed4 November 2026

Markdown ยท JSON ยท Edit on GitHub

What this page needs for W1

  • an inventory row: the row of the inventory that the node accounts for (0 of 1)

The writing levels say how complete the text of a page is. They do not measure its truth.

What links here

No page links to this page yet.