03.06.13 · modern-geometry / characteristic-classes

Oriented bordism and the Pontryagin-Thom construction

shipped3 tiersLean: partial

Anchor (Master): Thom 1954 *Comment. Math. Helv.* 28 *Quelques propriétés globales des variétés différentiables* (oriented case in the same paper as the unoriented); Wall 1960 *Ann. Math.* 72, 292-311 *Determination of the cobordism ring* (integral computation of Ω^SO_* up to dimension 8); Milnor-Stasheff §17 (Pontryagin-Thom collapse for MSO); Stong *Notes on Cobordism Theory* (Princeton, 1968) Ch. V; Novikov 1962 *Izv. Akad. Nauk SSSR* 26 (complex cobordism)

Intuition [Beginner]

Oriented bordism asks the same geometric question as unoriented bordism, but with one extra constraint: every shape carries a choice of direction of rotation, and the bordism between two shapes must respect that direction. Two oriented circles are oriented-bordant when together — with one of them flipped to absorb a sign — they form the oriented edge of a higher-dimensional oriented region. The orientation discipline is exactly what distinguishes a Möbius band from a cylinder: the cylinder is oriented, the band is not.

When orientations are part of the data, the bordism group becomes much richer than in the unoriented case. Elements no longer all have order two; the bordism class of the complex projective plane has infinite order, and the resulting group rationally is a polynomial ring on the even-dimensional complex projective spaces. The cost is more arithmetic structure to keep track of. The reward is a sharper invariant that detects oriented manifolds modulo oriented boundaries, and that connects to integer-valued invariants like the signature.

The Pontryagin-Thom construction translates this geometric problem into a problem about homotopy groups of a specific spectrum. The unoriented version reads bordism off the Thom spectrum of the universal real vector bundle. The oriented version replaces this by the Thom spectrum of the universal oriented real vector bundle, called MSO, and the same translation occurs. Thom 1954 and Wall 1960 together completed the calculation. The slogan: a closed oriented manifold is an oriented boundary if and only if all its Pontryagin numbers vanish and certain Stiefel-Whitney numbers vanish.

Visual [Beginner]

Picture an annulus — a flat ring with two boundary circles — with arrows drawn along each circle. If both arrows go counterclockwise, the annulus is an oriented bordism between two copies of the standard oriented circle. If you reverse one arrow, the annulus is an oriented bordism from the standard circle to the opposite circle, equivalently from the sum to zero in the oriented bordism group.

Going up two dimensions, the complex projective plane carries a canonical orientation coming from its complex structure. The class of this manifold in the oriented bordism group has infinite order — no number of disjoint copies of it, however many, is the oriented boundary of any compact five-dimensional region. This is the fundamental difference between oriented and unoriented bordism: in the unoriented setting every class has order two, in the oriented setting the projective plane already has infinite order.

Worked example [Beginner]

The two-sphere $S^{2}$ with its standard orientation is bordant-with-orientation to the empty manifold. The bordism is the closed three-ball $B^{3}$ with its standard orientation, whose boundary is the oriented two-sphere.

Step 1. Equip the round two-sphere with the orientation coming from outward normal vectors of $B^{3}$ .

Step 2. Take the solid three-ball with its standard orientation as a subset of three-space.

Step 3. Check that the oriented boundary of the solid ball is the single oriented two-sphere with the standard outward orientation.

Step 4. Conclude that the oriented bordism class of the oriented $S^{2}$ equals the zero class in $Ω_{2}^{S O} = 0$ .

What this tells us: oriented two-manifolds are entirely classified by oriented bordism into a single zero element. The first substantive oriented bordism group sits in dimension four, generated by the complex projective plane.

Check your understanding [Beginner]

Exercise (easy, multiple choice).

Two closed oriented manifolds are called oriented-bordant when:

A. Their disjoint union is the unoriented boundary of a higher-dimensional manifold. B. There is a compact oriented manifold whose oriented boundary is one minus the other. C. They have the same dimension. D. They have the same orientation.

Hint

The oriented case differs from the unoriented case in that orientations must match up along the boundary, with appropriate sign.

Answer

B. Two oriented manifolds $M$ and $N$ are oriented-bordant when there is a compact oriented $W$ whose oriented boundary equals the disjoint union of $M$ with the orientation-reversed copy of $N$ , equivalently when the difference $[M] - [N]$ is the boundary class of $W$ in the oriented bordism group.

Feedback-correct: yes; the bar denotes orientation reversal. Feedback-wrong: same dimension is necessary but not sufficient; unoriented bordism is a strictly weaker relation.

Formal definition [Intermediate+]

Let $M_{n}^{S O}$ denote the class of closed smooth oriented $n$ -manifolds (compact, without boundary, equipped with an integral orientation class $[M] \in H_{n} (M; Z)$ ). Define the relation $M \sim_{S O} N$ on $M_{n}^{S O}$ by declaring $M \sim_{S O} N$ when there exists a compact smooth oriented $(n + 1)$ -manifold $W$ with oriented boundary $\partial W ≅ M ⊔ \overset{ˉ}{N}$ , where $\overset{ˉ}{N}$ denotes $N$ with reversed orientation. The relation $\sim_{S O}$ is an equivalence relation, with reflexivity from the cylinder $M \times [0, 1]$ , symmetry from the orientation-reversing diffeomorphism $W \to W$ that flips the bordism direction, and transitivity from the oriented collaring theorem.

Definition (oriented bordism group). The set $$ \Omega^{SO}_n := \mathcal{M}n^{SO} \big/ \sim{SO} $$ is an abelian group with addition $[M] + [N] = [M ⊔ N]$ , additive inverse $- [M] = [\overset{ˉ}{M}]$ (because $M ⊔ \overset{ˉ}{M} = \partial (M \times [0, 1])$ as oriented manifolds), and zero $[\emptyset]$ .

Definition (oriented bordism ring). The direct sum $$ \Omega^{SO}* := \bigoplus{n \geq 0} \Omega^{SO}_n $$ becomes a graded-commutative ring under Cartesian product $[M] \cdot [N] = [M \times N]$ , with the product orientation. The unit is the class of the positively oriented one-point space in dimension zero.

The Pontryagin classes $p_{i} (T M) \in H^{4 i} (M; Z)$ of an oriented $n$ -manifold (see 03.06.04) yield integral characteristic invariants. A Pontryagin number of a closed oriented $4 k$ -manifold $M$ is, for any partition $I = (i_{1}, \dots, i_{r})$ of $k$ with $i_{1} + \dots + i_{r} = k$ , the integer $$ p_I(M) := \langle p_{i_1}(TM) \cdot p_{i_2}(TM) \cdots p_{i_r}(TM),\ [M] \rangle \in \mathbb{Z}, $$ where $[M] \in H_{4 k} (M; Z)$ is the integral fundamental class. For dimensions not divisible by four the Pontryagin classes contribute no top-degree information; in those dimensions oriented Stiefel-Whitney numbers and integral Pontryagin numbers in lower degrees control the bordism class together.

Counterexamples to common slips [Intermediate+]

Oriented bordism is sharper than unoriented bordism. The forgetful map $Ω_{n}^{S O} \to N_{n}$ is generally neither injective nor surjective. The class of $CP^{2}$ is non-zero in $Ω_{4}^{S O}$ but maps to a non-zero element of $N_{4}$ that already corresponds to a different generator. The class of the Wu manifold in dimension five is a non-zero element of $Ω_{5}^{S O} ≅ Z /2$ whose image in $N_{5}$ might differ from the Dold-manifold generator. Confusing the two relations is the standard slip.
Orientation reversal versus complex conjugation. For complex manifolds, complex conjugation acts on $CP^{n}$ but does not reverse its standard orientation when $n$ is even — because $CP^{n}$ has real dimension $2 n$ , and complex conjugation acts on the tangent space as an $R$ -linear map whose determinant has sign $(- 1)^{n}$ . So $CP^{2 k}$ is not orientation-reversible by conjugation alone, which is one reason it generates a $Z$ summand. Confusing complex-conjugation symmetry with orientation reversal is a common error.
Integral fundamental class versus mod-2. In oriented bordism the integral fundamental class $[M] \in H_{n} (M; Z)$ is required; in unoriented bordism only the mod-2 class is needed. Pontryagin numbers pair with the integral class and live in $Z$ ; reducing the integral class mod 2 collapses Pontryagin numbers to expressions involving Stiefel-Whitney classes via Wu's formula.

Key theorem with proof [Intermediate+]

Theorem (Bordism invariance of Pontryagin numbers). If $M$ and $N$ are oriented-bordant closed smooth $4 k$ -manifolds, then for every partition $I$ of $k$ the Pontryagin numbers agree as integers: $$ p_I(M) = p_I(N) \in \mathbb{Z}. $$

Proof. Let $W$ be a compact smooth oriented $(4 k + 1)$ -manifold with oriented boundary $\partial W = M ⊔ \overset{ˉ}{N}$ , and let $i_{M} : M ↪ W$ and $i_{N} : N ↪ W$ be the oriented boundary inclusions. The tangent bundle of $W$ , restricted along the boundary inclusion, splits as $$ i_M^* TW \cong TM \oplus \underline{\mathbb{R}}, $$ where $\underline{R}$ is the product line bundle spanned by the inward-pointing normal vector. Pontryagin classes are stable under direct sum with a product bundle (03.06.04), so $$ i_M^* p_i(TW) = p_i(TM) $$ for every $i \geq 1$ . The same identity holds for $i_{N}$ .

Fix a partition $I = (i_{1}, \dots, i_{r})$ with $∣ I ∣ := \sum i_{j} = k$ . Set $$ W_I := p_{i_1}(TW) \cdots p_{i_r}(TW) \in H^{4k}(W; \mathbb{Z}). $$ Then $i_{M}^{*} W_{I} = p_{i_{1}} (T M) \dots p_{i_{r}} (T M) =: M_{I}$ , and similarly $i_{N}^{*} W_{I} = N_{I}$ .

The oriented boundary relation in integral homology reads $$ [M] - [N] = \partial [W] \quad \text{in } H_{4k}(W; \mathbb{Z}), $$ where $[W] \in H_{4 k + 1} (W, \partial W; Z)$ is the relative integral fundamental class of the manifold-with-boundary, and the boundary map of the long exact sequence of the pair sends $[W]$ to $[\partial W] = [M] - [N]$ inside $H_{4 k} (\partial W; Z) ≅ H_{4 k} (M; Z) \oplus H_{4 k} (N; Z)$ . The sign appears because the oriented boundary of $W$ is $M$ disjoint with $N$ -with-reversed-orientation.

Pair this with $W_{I}$ pulled back to the boundary: $$ \langle W_I, [M] \rangle - \langle W_I, [N] \rangle = \langle W_I, [M] - [N] \rangle = \langle W_I, \partial [W] \rangle. $$ By the relative-pairing identity (Stokes in integral cohomology), $$ \langle W_I, \partial [W] \rangle = \langle \delta W_I, [W] \rangle, $$ and $δ W_{I} = 0$ since $W_{I}$ is already a cocycle: Pontryagin classes are pulled back from the classifying space $B S O$ and are absolute cohomology classes, not relative ones. Therefore $$ \langle M_I, [M] \rangle - \langle N_I, [N] \rangle = 0 \in \mathbb{Z}, $$ i.e. $p_{I} (M) = p_{I} (N)$ . $□$

Bridge. The Pontryagin-number bordism invariance is the foundational reason that $Ω_{*}^{S O}$ admits a well-defined integer-valued character for each partition; this is exactly the calculation Wall 1960 promotes — together with the Stiefel-Whitney-number companion — into a complete invariant. The argument generalises the unoriented template from 03.06.12: appears again as the rational separation result $Ω_{*}^{S O} \otimes Q ↪ \prod_{k} Hom (P_{k}, Q)$ via Pontryagin-number characters, and builds toward the Pontryagin-Thom construction, which identifies $Ω_{*}^{S O}$ with $π_{*} (M S O)$ . The central insight is that any characteristic-class evaluation against the integral fundamental class is automatically an oriented bordism invariant — a stably abelian shadow of Stokes' theorem in integer coefficients.

Exercises [Intermediate+]

Exercise 3 (medium, short answer).

Explain why $Ω_{n}^{S O}$ is not an $F_{2}$ -vector space in general, in contrast with $N_{n}$ .

Hint

In unoriented bordism, $M ⊔ M = \partial (M \times [0, 1])$ . In oriented bordism, the same cylinder has oriented boundary $M ⊔ \overset{ˉ}{M}$ , not $M ⊔ M$ .

Answer

The cylinder $M \times [0, 1]$ has oriented boundary $M ⊔ \overset{ˉ}{M}$ in the oriented category, so it witnesses $[M] + [- M] = 0$ — equivalent to the existence of an additive inverse via orientation reversal, not 2-torsion. For 2-torsion we would need $M ⊔ M$ to be an oriented boundary, which happens only if $M$ admits an orientation-reversing diffeomorphism (in which case $M ≅ \overset{ˉ}{M}$ as oriented manifolds, forcing $2 [M] = [M] + [\overset{ˉ}{M}] = 0$ ). Manifolds like $CP^{2}$ have no orientation-reversing diffeomorphism — the signature provides an obstruction — so $[CP^{2}]$ has infinite order.

Exercise 4 (medium, symbolic).

Compute the total Pontryagin class of $T (CP^{n})$ in terms of the hyperplane class $H \in H^{2} (CP^{n}; Z)$ , and use it to find $p_{1}$ and $p_{2}$ for $T (CP^{4})$ .

Hint

The Euler sequence gives $T C P^{n} \oplus \underline{C} ≅ (n + 1) O (1)$ over $C$ . Use the relation between Pontryagin and Chern classes: $p_{i} (E_{R}) = (- 1)^{i} c_{2 i} (E \oplus \overset{ˉ}{E}) = \dots$ and the standard formula $p (E_{R}) = c (E) \cdot c (\overset{ˉ}{E})$ for the real form of a complex bundle.

Answer

The total Chern class of $T CP^{n}$ is $c (T CP^{n}) = (1 + H)^{n + 1}$ , and Pontryagin classes of the underlying real bundle satisfy $1 - p_{1} + p_{2} - \dots = c (T) \cdot c (\overset{ˉ}{T}) = (1 + H)^{n + 1} \cdot (1 - H)^{n + 1} = (1 - H^{2})^{n + 1}$ . Expanding: $$ p(T\mathbb{CP}^n) = (1 + H^2)^{n+1} $$ (absorbing the sign by working with $p_{i}$ as the unsigned coefficients). For $n = 4$ this gives $p (T CP^{4}) = (1 + H^{2})^{5} = 1 + 5 H^{2} + 10 H^{4} + 10 H^{6} + 5 H^{8} + H^{10}$ . Since $H^{5} = 0$ in $CP^{4}$ , we read off $p_{1} (T CP^{4}) = 5 H^{2}$ and $p_{2} (T CP^{4}) = 10 H^{4}$ .

Exercise 6 (medium, short answer).

State the splitting that the forgetful map $Ω_{n}^{S O} \to N_{n}$ admits rationally, and use it to deduce that $Ω_{n}^{S O} \otimes Q = 0$ when $n$ is not divisible by $4$ .

Hint

Pontryagin numbers separate $Ω_{*}^{S O} \otimes Q$ . Top-degree Pontryagin monomials only exist in dimensions divisible by $4$ .

Answer

The Pontryagin-number map $Ω_{n}^{S O} \otimes Q \to \prod_{∣ I ∣ = n /4} Q$ defined by $[M] \mapsto (p_{I} (M))_{I}$ is injective by Thom-Wall's rational computation $Ω_{*}^{S O} \otimes Q ≅ Q [CP^{2}, CP^{4}, \dots]$ . In dimensions $n$ not divisible by $4$ , there are no length- $n$ partitions from multiples of $4$ , hence no rational Pontryagin numbers; the rational bordism group vanishes. Concretely $Ω_{1}^{S O} = Ω_{2}^{S O} = Ω_{3}^{S O} = 0$ already integrally; $Ω_{5}^{S O} ≅ Z /2$ is pure torsion and rationalises to zero; $Ω_{6}^{S O} = Ω_{7}^{S O} = 0$ again.

Exercise 7 (hard, short answer).

State the Pontryagin-Thom collapse map in the oriented case for a smooth oriented embedding $f : M^{n} ↪ S^{n + k}$ and explain why its homotopy class lives in $π_{n + k} (M S O (k))$ rather than $π_{n + k} (M O (k))$ .

Hint

The normal bundle of an oriented manifold embedded in a sphere inherits an orientation from the embedding plus the ambient orientation. This refines the classifying map.

Answer

The tubular-neighbourhood theorem gives a smooth open neighbourhood $U \subset S^{n + k}$ of $f (M)$ diffeomorphic to the open disk bundle of the normal bundle $ν_{f}$ . Because the ambient sphere and the embedded $M$ are oriented, the normal bundle $ν_{f}$ inherits a canonical orientation: the orientation of $S^{n + k}$ decomposes as the orientation of $M$ times the orientation of $ν_{f}$ at every point. The Pontryagin-Thom collapse $c : S^{n + k} \to Th (ν_{f})$ sends the closed disk bundle to the corresponding Thom-space disc and the exterior to the base point. Composing with the oriented classifying map $Th (ν_{f}) \to Th (γ_{S O (k)}^{k}) = M S O (k)$ — which exists because $ν_{f}$ has structure group $S O (k)$ — yields a map $S^{n + k} \to M S O (k)$ , hence an element of $π_{n + k} (M S O (k))$ . The lift through $M S O$ rather than $M O$ records the orientation data: stripping the orientation gives the map composed with $M S O \to M O$ and recovers the unoriented Pontryagin-Thom class.

Exercise 8 (hard, short answer).

Identify the rational polynomial generators of $Ω_{*}^{S O} \otimes Q$ as a graded $Q$ -algebra. Which dimensions admit a generator, and which manifolds realise them?

Hint

Rational generators exist exactly in dimensions divisible by four. The even-dimensional complex projective spaces $CP^{2 k}$ provide them.

Answer

Thom 1954 / Wall 1960: $Ω_{*}^{S O} \otimes Q ≅ Q [CP^{2}, CP^{4}, CP^{6}, \dots]$ as a graded $Q$ -algebra, with one polynomial generator $[CP^{2 k}]$ in each dimension $4 k$ for $k \geq 1$ . Hence:

$dim_{Q} (Ω_{0}^{S O} \otimes Q) = 1$ , $dim_{Q} (Ω_{4}^{S O} \otimes Q) = 1$ generated by $[CP^{2}]$ ,
$dim_{Q} (Ω_{8}^{S O} \otimes Q) = 2$ generated by $[CP^{4}]$ and $[CP^{2}]^{2} = [CP^{2} \times CP^{2}]$ ,
$dim_{Q} (Ω_{12}^{S O} \otimes Q) = 3$ generated by $[CP^{6}]$ , $[CP^{2}] [CP^{4}]$ , $[CP^{2}]^{3}$ ,
in general $dim_{Q} (Ω_{4 k}^{S O} \otimes Q) = p (k)$ , the number of partitions of $k$ , with basis given by monomials in $[CP^{2 k_{1}}] \dots [CP^{2 k_{r}}]$ over partitions $(k_{1}, \dots, k_{r})$ of $k$ . In odd-multiple-of-4 dimensions (i.e. $n$ not divisible by $4$ ), the rational bordism group is zero.

Exercise 9 (hard, short answer).

Compute $Ω_{4}^{S O}$ integrally and describe the signature homomorphism $σ : Ω_{4}^{S O} \to Z$ .

Hint

Wall 1960: $Ω_{4}^{S O} ≅ Z$ generated by $[CP^{2}]$ . The signature equals one on this generator.

Answer

Wall 1960 Ann. Math. 72 computed $Ω_{4}^{S O} ≅ Z$ , with generator $[CP^{2}]$ . The signature $σ : Ω_{4}^{S O} \to Z$ defined by the intersection form on middle cohomology is a ring homomorphism — in fact the same map as the rational Hirzebruch $L$ -genus at degree four, but already integer-valued. On the generator, $σ (CP^{2}) = 1$ , witnessed by the intersection form $⟨ 1 ⟩$ on $H^{2} (CP^{2}; Z) = Z ⟨ H ⟩$ . Therefore $σ$ is an isomorphism $Ω_{4}^{S O} \sim Z$ , and any closed oriented four-manifold $M$ is oriented-bordant to $σ (M)$ disjoint copies of $CP^{2}$ (or $- σ (M)$ copies of $\overset{ˉ}{CP}^{2}$ if the signature is negative). Hirzebruch's formula $3 σ (M) = p_{(1)} (M)$ in dimension four says the signature equals one third of the first Pontryagin number.

Exercise 10 (hard, short answer).

State Thom's main theorem $Ω_{*}^{S O} ≅ π_{*} (M S O)$ and explain why the right-hand side rationally is the polynomial ring $Q [CP^{2}, CP^{4}, \dots]$ via the rational equivalence $M S O_{Q} ≃ ⋁_{k} H Q$ on appropriate skeleta.

Hint

Rationally the Thom spectrum splits as a wedge of rational Eilenberg-MacLane spectra in degrees divisible by four. Pontryagin numbers correspond to the cohomology classes of the wedge.

Answer

Thom 1954 / Milnor-Stasheff §17: the oriented Pontryagin-Thom collapse construction induces a ring isomorphism $$ \Phi : \Omega^{SO}* \xrightarrow{\sim} \pi(MSO), $$ where $M S O = colim_{k} Σ^{- k} Th (γ_{S O (k)}^{k})$ is the oriented Thom spectrum. The right-hand side computes rationally via the Adams spectral sequence with $E_{2}$ -page $\mathrm{Ext}_{\mathcal{A}}^{,}(H^(MSO; \mathbb{F}2), \mathbb{F}2) $(a tt h e p r im e tw o) or v ia t h er a t i o na l P os t nik o v d eco m p os i t i o n (r a t i o na l l y) . R a t i o na l l y,$ H^(MSO; \mathbb{Q}) $i s a f r ee g r a d e d$ \mathbb{Q} $- m o d u l eco n ce n t r a t e d in d e g r ees d i v i s ib l e b y f o u r, w i t hba s i s in d e x e d b y p a r t i t i o n s; t hi sco m es f r o m t h e P o n t r y a g in c l a sseso f t h e u ni v er s a l or i e n t e d b u n d l e, s in ce$ H^(BSO; \mathbb{Q}) = \mathbb{Q}[p_1, p_2, \ldots] $i s a p o l y n o mia l r in g o n t h e u ni v er s a l P o n t r y a g in c l a sses . B y t h er a t i o na l s pl i tt in g p r in c i pl e an d H a tt or i - S t o n g, t h er a t i o na l h o m o t o p y o f$ MSO $i se x a c tl y i t sr a t i o na l co h o m o l o g y r e a d o f f t h e T h o m c l a ss, y i e l d in g$ \pi*(MSO) \otimes \mathbb{Q} \cong \mathbb{Q}[a_4, a_8, a{12}, \ldots] $. C o mbin e d w i t h$ \Phi $, t h e i d e n t i f i c a t i o ni s$ \Omega^{SO}* \otimes \mathbb{Q} \cong \mathbb{Q}[\mathbb{CP}^2, \mathbb{CP}^4, \ldots] $a s g r a d e d r in g s, w i t h$ |\mathbb{CP}^{2k}| = 4k $. I n t e g r a l l y, t or s i o n s u b g r o u p so f$ \pi(MSO) $a r eco n ce n t r a t e d a tt h e p r im e 2 (N o v ik o v 1962, * M a t . S b . *) an d ma t c hW a l l^{'} s 1960 co m p u t a t i o n o f$ \Omega^{SO}_$ integrally through dimension 8.

Lean formalization [Intermediate+]

lean_status: partial — the module Codex.Modern.CharClasses.OrientedBordismPT declares the central definitions and the bordism-invariance lemma as theorems with sorry-bodies, plus decidable numerical witnesses for the Hirzebruch identities at $CP^{2}$ and $CP^{4}$ .

-- File: lean/Codex/Modern/CharClasses/OrientedBordismPT.lean

structure ClosedSmoothOrientedManifold (n : ℕ) where
  carrier : Type
  isClosed : True
  isOriented : True

def IsOrientedBordant {n : ℕ} (_M _N : ClosedSmoothOrientedManifold n) : Prop := True

theorem isOrientedBordant_refl  : ∀ {n} M, IsOrientedBordant (n := n) M M := …
theorem isOrientedBordant_symm  : …
theorem isOrientedBordant_trans : …

def OrientedBordismGroup (n : ℕ) : Type := Unit
instance : AddCommGroup (OrientedBordismGroup n) := …

-- Central lemma: bordism invariance of Pontryagin numbers.
theorem pontryagin_number_bordism_invariant {n : ℕ}
    {M N : ClosedSmoothOrientedManifold n}
    (_h : IsOrientedBordant M N) (I : List ℕ) :
    pontryaginNumber M I = pontryaginNumber N I := …

-- Thom's main theorem (statement).
theorem thom_oriented_main_theorem (n : ℕ) :
    OrientedBordismGroup n ≃+ MSOHomotopy n := …

-- Hirzebruch numerical witnesses (decidable).
def sig_CP2 : ℤ := 1
def p1_CP2  : ℤ := 3
theorem hirzebruch_at_CP2 : 3 * sig_CP2 = p1_CP2 := by decide

def sig_CP4   : ℤ := 1
def p2_CP4    : ℤ := 10
def p1sq_CP4  : ℤ := 25
theorem hirzebruch_at_CP4 : 45 * sig_CP4 = 7 * p2_CP4 - p1sq_CP4 := by decide

The full proof bodies depend on three Mathlib gaps: (1) smooth oriented manifolds with boundary and the oriented collaring theorem; (2) the Pontryagin class API on oriented real vector bundles, with integral fundamental classes for closed oriented manifolds; (3) the stable-homotopy category, the Thom spectrum $M S O$ , and the rational equivalence $M S O_{Q} ≃ ⋁_{k} H Q$ . The Lean module is shipped as a target API, not a finished formalisation. The numerical witnesses $σ (CP^{2}) = 1$ with $p_{1} = 3$ , and $σ (CP^{4}) = 1$ with $p_{2} = 10$ , $p_{1}^{2} = 25$ , are recorded as decide-provable identities at the integer level.

Advanced results [Master]

The oriented bordism ring $Ω_{*}^{S O}$

Pontryagin (1947, 1950) introduced the bordism relation in its framed form to compute homotopy groups of spheres ^[Rohlin]. Thom 1954 ^[Thom] generalised the construction to arbitrary structure groups; in the same paper that founds unoriented bordism via $M O$ , Thom introduced the oriented Thom spectrum $M S O = colim_{k} Σ^{- k} Th (γ_{S O (k)}^{k})$ and stated the analogous identification $Ω_{n}^{S O} ≅ π_{n} (M S O)$ .

The oriented bordism group $Ω_{n}^{S O}$ has more structure than its unoriented counterpart. Disjoint union $[M] + [N] = [M ⊔ N]$ defines the abelian-group operation. Orientation reversal $[M] \mapsto [\overset{ˉ}{M}]$ provides additive inverses: the cylinder $M \times [0, 1]$ has oriented boundary $M ⊔ \overset{ˉ}{M}$ , hence $[M] + [\overset{ˉ}{M}] = 0$ . Unlike $N_{n}$ , elements of $Ω_{n}^{S O}$ need not be 2-torsion; the obstruction is precisely whether $M$ admits an orientation-reversing diffeomorphism. The ring structure $[M] \cdot [N] = [M \times N]$ uses the product orientation; graded-commutativity is signed, with $[M] \cdot [N] = (- 1)^{d i m M \cdot d i m N} [N] \cdot [M]$ after careful orientation tracking, though the ring is in fact strictly commutative on the level of bordism classes because reversing two factor orientations restores the original orientation.

The first low-dimensional values, integrally ^[Wall]: $$ \Omega^{SO}_0 = \mathbb{Z}, \quad \Omega^{SO}_1 = \Omega^{SO}_2 = \Omega^{SO}_3 = 0, $$ $$ \Omega^{SO}_4 = \mathbb{Z} \langle [\mathbb{CP}^2] \rangle, \quad \Omega^{SO}_5 = \mathbb{Z}/2, \quad \Omega^{SO}_6 = \Omega^{SO}_7 = 0, $$ $$ \Omega^{SO}_8 = \mathbb{Z} \langle [\mathbb{CP}^4] \rangle \oplus \mathbb{Z} \langle [\mathbb{CP}^2 \times \mathbb{CP}^2] \rangle. $$ The dimension-five $Z /2$ is represented by the Wu manifold $W = S U (3) / S O (3)$ — a five-dimensional homogeneous space whose Stiefel-Whitney number $w_{2} w_{3}$ does not vanish; it is the lowest-dimensional torsion class in oriented bordism. The pattern is governed by the rational computation stated below, together with the small-prime Adams spectral sequence (Novikov 1962 ^[Novikov]).

Pontryagin-Thom via the MSO Thom spectrum

Let $γ_{S O (k)}^{k} \to B S O (k)$ be the universal oriented rank- $k$ real vector bundle. Its Thom space is $Th (γ_{S O (k)}^{k}) := D (γ_{S O (k)}^{k}) / S (γ_{S O (k)}^{k})$ where $D, S$ are the closed disk and sphere bundles. Define $M S O (k) := Th (γ_{S O (k)}^{k})$ . The structure maps $Σ M S O (k) \to M S O (k + 1)$ — coming from the inclusion $γ_{S O (k)}^{k} \oplus \underline{R} ↪ γ_{S O (k + 1)}^{k + 1} ∣_{B S O (k)}$ — assemble these spaces into a spectrum, the oriented Thom spectrum $M S O$ ^{[Milnor-Stasheff §17]}.

The Pontryagin-Thom construction in the oriented setting takes a closed smooth oriented $n$ -manifold $M$ , embeds it into a sphere $S^{n + k}$ for $k$ large (Whitney's embedding theorem in the oriented category, which proceeds with no extra structure), and takes the normal bundle $ν_{M}$ . The normal bundle inherits an orientation from the orientations of $M$ and the ambient sphere; concretely, at each point $p \in M$ , the orientation of $T_{p} S^{n + k}$ is the product of the orientation of $T_{p} M$ and the orientation of $(ν_{M})_{p}$ , and this rule globalises to an orientation of $ν_{M}$ .

The collapse map $$ c_M : S^{n+k} \longrightarrow \mathrm{Th}(\nu_M) $$ sends the disk bundle of $ν_{M}$ by the identity and the complement to the base point. Composing $c_{M}$ with the oriented classifying map $Th (ν_{M}) \to M S O (k)$ — built from a bundle map $ν_{M} \to γ_{S O (k)}^{k}$ that respects orientations — yields a map $S^{n + k} \to M S O (k)$ , i.e. an element of $π_{n + k} (M S O (k))$ . Passing to the stable colimit gives the oriented Thom homomorphism $$ \Phi^{SO} : \Omega^{SO}_n \longrightarrow \pi_n(MSO). $$

Three properties of $Φ^{S O}$ require checking, all standard from Milnor-Stasheff §17 ^{[Milnor-Stasheff]}: (i) well-definedness modulo embedding, using the regular-homotopy theorem in the oriented category; (ii) zero on boundaries, using the relative tubular neighbourhood of an oriented null-bordism $W$ embedded in $D^{n + k + 1}$ ; (iii) multiplicativity, with $c_{M \times N} = c_{M} \land c_{N}$ on smash-product spectra. The orientation discipline ripples through each step: the regular-homotopy and tubular-neighbourhood arguments are unchanged, but the classifying maps must factor through $B S O (k)$ rather than $B O (k)$ .

Thom's main theorem $Ω_{}^{S O} ≅ π_{} (M S O)$ and the rational computation

Theorem (Thom 1954 / Wall 1960). The oriented Thom homomorphism $\Phi^{SO} : \Omega^{SO}_ \to \pi_*(MSO)$ is an isomorphism of graded rings.*

Injectivity comes from oriented transversality: a null-homotopy of the oriented collapse map produces, by transversality with the zero section of $γ_{S O (k)}^{k}$ in the oriented category, an oriented bordism $W$ between $M$ and the empty manifold. Surjectivity uses the same transversality argument applied to an arbitrary stable map $S^{n + k} \to M S O (k)$ : making the map transverse to $B S O (k) ↪ M S O (k)$ produces a closed oriented manifold of dimension $n$ whose oriented Pontryagin-Thom class is the given homotopy class. The orientation of the preimage manifold comes from the orientation of the target Thom space.

The right-hand side $π_{*} (M S O)$ is computed in two flavours. Rationally, $H^{*} (M S O; Q)$ is concentrated in degrees divisible by four, with $H^{*} (B S O; Q) = Q [p_{1}, p_{2}, \dots]$ the polynomial ring on the universal Pontryagin classes; the Thom isomorphism $H^{*} (M S O; Q) = H^{*} (B S O; Q) \cdot U$ then gives a free polynomial structure indexed by partitions. By the rational Hurewicz theorem applied stably, $π_{*} (M S O) \otimes Q$ matches its rational cohomology after dualisation. Pontryagin numbers serve as the indexing characters. The result: $$ \pi_*(MSO) \otimes \mathbb{Q} \cong \mathbb{Q}[a_4, a_8, a_{12}, \ldots] $$ as a graded $Q$ -algebra, with $∣ a_{4 k} ∣ = 4 k$ .

The geometric realisation of the generators is the central calculation: $[CP^{2 k}] \in Ω_{4 k}^{S O}$ provides a rational polynomial generator in dimension $4 k$ for each $k \geq 1$ . Equivalently: $$ \Omega^{SO}* \otimes \mathbb{Q} \cong \mathbb{Q}[\mathbb{CP}^2, \mathbb{CP}^4, \mathbb{CP}^6, \ldots] $$ with $∣ CP^{2 k} ∣ = 4 k$ . The proof that ${[\mathbb{CP}^{2k}]}{k \geq 1} $i s a l g e b r ai c a l l y in d e p e n d e n t u ses P o n t r y a g inn u mb er s : t h e P o n t r y a g in - n u mb er ma t r i x o n t h e p a r t i t i o nba s i s i s u pp er t r ian g u l a r w i t h p os i t i v e d ia g o na l e n t r i es (H i r z e b r u c h^{'} s$ L$-genus identity at the diagonal), hence invertible rationally ^[Hirzebruch].

Integrally, $π_{*} (M S O)$ has 2-torsion that Wall 1960 ^[Wall] worked out completely up to dimension 8. The integral structure is more intricate: $Ω_{5}^{S O} = Z /2$ from the Wu manifold; in higher dimensions Wall identified all torsion as 2-primary, with explicit generators. The Adams spectral sequence at the prime 2 (Novikov 1962 ^[Novikov]) systematises this.

Wall 1960 — Pontryagin and Stiefel-Whitney numbers classify $Ω_{*}^{S O}$

The unoriented theorem (Thom 1954) said: Stiefel-Whitney numbers separate $N_ $. * T h eor i e n t e d r e f in e m e n t (W a l l 1960 [r e f : T O D O_{R} E F W a l l]) s a y s : * P o n t r y a g inn u mb er s an d S t i e f e l - W hi t n ey n u mb er s t o g e t h er se p a r a t e$ \Omega^{SO}_$. Concretely, the map $$ (\mathrm{Pont}, \mathrm{SW}) : \Omega^{SO}_n \longrightarrow \mathrm{Hom}(\mathcal{P}^{\mathrm{Pont}}_n, \mathbb{Z}) \oplus \mathrm{Hom}(\mathcal{P}^{\mathrm{SW}}_n, \mathbb{F}_2) $$ is injective, where $P_{n}^{Pont}$ is the partition basis indexing top-degree Pontryagin monomials (non-empty only when $4∣ n$ ) and $P_{n}^{SW}$ is the partition basis indexing top-degree Stiefel-Whitney monomials.

Rationally, Pontryagin numbers alone suffice: $Ω_{*}^{S O} \otimes Q ↪ \prod_{n} Hom (P_{n}^{Pont}, Q)$ via the Pontryagin-number characters, and this map is in fact an injection of rational vector spaces of equal dimension (both sides have dimension $p (n /4)$ , the number of partitions of $n /4$ ), hence an isomorphism by counting. The Hirzebruch signature theorem 03.06.11 is the most famous specialisation: $σ : Ω_{*}^{S O} \to Z$ is the ring homomorphism corresponding to the $L$ -genus, killing torsion and computing $σ (CP^{2 k}) = 1$ on every generator.

The Stiefel-Whitney numbers are needed precisely for the 2-torsion: the Wu manifold has all Pontryagin numbers zero (the dimension is not divisible by four) but a non-zero Stiefel-Whitney number $w_{2} w_{3}$ , witnessing the $Z /2$ -summand in $Ω_{5}^{S O}$ . More generally, the 2-torsion of $Ω_{*}^{S O}$ is detected by Stiefel-Whitney numbers via the mod-2 reduction map $Ω_{*}^{S O} \to N_{*}$ combined with the unoriented Thom-Stiefel-Whitney machine.

Synthesis. Thom's theorem for oriented bordism is the foundational reason that the rational ring $Ω_{*}^{S O} \otimes Q$ is a polynomial $Q$ -algebra on the even-dimensional complex projective spaces — a remarkably clean closed form for a geometric invariant. The central insight is the oriented Pontryagin-Thom collapse: an oriented manifold embedded in a sphere with oriented normal bundle becomes a map from a high sphere into $M S O (k)$ , and the entire oriented bordism class is captured by the stable homotopy class of that map. Putting these together with Wall 1960's integral computation through dimension eight, the oriented bordism ring is identified with the homotopy ring of $M S O$ — rationally a clean polynomial ring on the $CP^{2 k}$ , integrally a graded ring with $2$ -primary torsion controlled by the Adams spectral sequence. This is exactly the route Hirzebruch 1956 ^[Hirzebruch] generalises into the signature theorem and the multiplicative-sequence formalism, and Novikov 1962 ^[Novikov] extends to complex cobordism via $M U$ .

The bridge is between three computations: bordism invariance of Pontryagin numbers (Stokes in integral cohomology); the oriented Pontryagin-Thom collapse (oriented transversality + tubular neighbourhood); and the rational Postnikov-Hurewicz decomposition of $M S O$ . This pattern recurs in every cobordism theory: complex $Ω_{*}^{U} ≅ π_{*} (M U)$ via Quillen 1969's formal-group-law identification, spin $Ω_{*}^{Spin}$ via $M Spin$ with KO-theoretic refinements through the Atiyah-Bott-Shapiro index map, framed $Ω_{*}^{fr} ≅ π_{*}^{s}$ via the sphere spectrum (Pontryagin's original 1947 case). Each computation generalises the same machine with a different universal bundle.

Full proof set [Master]

Proposition (Oriented bordism is an equivalence relation). The relation $M \sim_{S O} N$ defined by the existence of $W$ with oriented boundary $M ⊔ \overset{ˉ}{N}$ is reflexive, symmetric, and transitive on closed smooth oriented $n$ -manifolds.

Proof. Reflexivity: $W = M \times [0, 1]$ has oriented boundary $\partial W = M ⊔ \overset{ˉ}{M}$ , so taking $N = M$ in the relation we use the rule that $\overset{ˉ}{M} ≅ M$ as smooth manifolds with the orientation reversed; the cylinder thus witnesses $M \sim_{S O} M$ . Strictly, we need $M \sim_{S O} M$ via the boundary $M ⊔ \overset{ˉ}{M}$ , which is the desired form when we replace the second factor by $\overset{ˉ}{M}$ in the relation definition.

Symmetry: a witness $W$ with $\partial W = M ⊔ \overset{ˉ}{N}$ is converted to a witness $\overset{ˉ}{W}$ with $\partial \overset{ˉ}{W} = \overset{ˉ}{M} ⊔ N$ , hence to a witness for $N \sim_{S O} M$ after reversing the order of summands; alternatively, take the same $W$ and use that the oriented boundary $\partial W = M ⊔ \overset{ˉ}{N}$ can be re-read as $N ⊔ \overset{ˉ}{M}$ after orientation reversal of all sides.

Transitivity: given $W_{1}$ with $\partial W_{1} = M ⊔ \overset{ˉ}{N}$ and $W_{2}$ with $\partial W_{2} = N ⊔ \overset{ˉ}{P}$ , glue along the common boundary component $N$ using the oriented collaring theorem (Milnor 1965 Topology from the differentiable viewpoint, oriented version). The collaring theorem provides oriented open collars $N \times [0, ϵ) \subset W_{1}$ and $N \times [0, ϵ) \subset W_{2}$ with the boundary embedded at $t = 0$ in each; identifying these collars with matching orientations produces a smooth oriented structure on $W_{1} \cup_{N} W_{2}$ whose oriented boundary is $M ⊔ \overset{ˉ}{P}$ . $□$

Proposition (Additive inverse via orientation reversal). For every closed smooth oriented $n$ -manifold $M$ , $[M] + [\overset{ˉ}{M}] = 0$ in $Ω_{n}^{S O}$ .

Proof. The cylinder $W = M \times [0, 1]$ has oriented boundary $\partial W = M ⊔ \overset{ˉ}{M}$ when we choose the standard product orientation and outward-normal convention. Therefore $[M] + [\overset{ˉ}{M}] = [M ⊔ \overset{ˉ}{M}] = [\partial W] = 0$ in $Ω_{n}^{S O}$ . The additive inverse of $[M]$ is $[\overset{ˉ}{M}]$ . $□$

Proposition (Bordism invariance of Pontryagin numbers). Established in the Key Theorem section.

Proposition (Multiplicativity of the oriented bordism ring). The product $[M] \cdot [N] := [M \times N]$ defines a well-defined commutative graded ring structure on $\Omega^{SO}_ $w i t h u ni t$ [\mathrm{pt}] \in \Omega^{SO}_0$.*

Proof. Well-definedness: if $M^{'} \sim_{S O} M$ via $W$ with $\partial W = M^{'} ⊔ \overset{ˉ}{M}$ , then $W \times N$ is an oriented cobordism between $M^{'} \times N$ and $M \times N$ — the boundary $\partial (W \times N) = \partial W \times N = (M^{'} ⊔ \overset{ˉ}{M}) \times N = (M^{'} \times N) ⊔ (\overset{ˉ}{M} \times N)$ , using that $N$ is closed so $\partial (W \times N) = (\partial W) \times N$ , and that $\overset{ˉ}{M} \times N = \overline{M \times N}$ as oriented manifolds when $N$ has even codimension contribution to the joint orientation. (For the strict orientation tracking, take both $M$ and $N$ even-dimensional or absorb the signs into the orientation-reversal convention; the resulting bordism still equates $[M^{'} \times N]$ with $[M \times N]$ in $Ω_{m + n}^{S O}$ .) Commutativity at the level of bordism classes: $M \times N ≅ N \times M$ via the swap diffeomorphism, which preserves the product orientation up to the sign $(- 1)^{mn}$ ; reversing one factor's orientation if necessary absorbs the sign, so $[M \times N] = [N \times M]$ in $Ω_{m + n}^{S O}$ . Associativity: $(M \times N) \times P ≅ M \times (N \times P)$ as oriented manifolds. Distributivity: $M \times (N ⊔ N^{'}) = (M \times N) ⊔ (M \times N^{'})$ at the level of oriented manifolds. The unit follows from $pt \times M = M$ as oriented smooth manifolds. $□$

Proposition (Oriented Pontryagin-Thom collapse is well-defined modulo oriented bordism). Let $f_{0}, f_{1} : M ↪ S^{n + k}$ be two smooth oriented embeddings of a closed smooth oriented $n$ -manifold. For $k$ large enough, the oriented collapse maps $c_{f_{0}}, c_{f_{1}} : S^{n + k} \to M S O (k)$ are stably homotopic. More generally, if $W$ is a compact oriented $(n + 1)$ -manifold with oriented boundary $\partial W = M_{0} ⊔ \overset{ˉ}{M}_{1}$ , and $F : W ↪ S^{n + k} \times [0, 1]$ extends $f_{0} ⊔ f_{1}$ on the boundary, then the oriented collapse maps differ by a homotopy.

Proof. The relative tubular-neighbourhood theorem in the oriented category gives an open tubular neighbourhood $V \subset S^{n + k} \times [0, 1]$ of $F (W)$ diffeomorphic to the oriented normal bundle $ν_{F}$ , and this restricts to an oriented tubular neighbourhood of $f_{i} (M_{i})$ on the slice $S^{n + k} \times {i}$ . The orientation of $ν_{F}$ extends the orientations of $ν_{f_{i}}$ at the two endpoints. The relative collapse map $$ S^{n+k} \times [0,1] \longrightarrow \mathrm{Th}(\nu_F) $$ quotients the complement of $V$ to the base point; composing with the oriented classifying map $Th (ν_{F}) \to M S O (k)$ produces a homotopy from $c_{f_{0}}$ to $c_{f_{1}}$ in the slice direction. Two oriented embeddings of the same $M$ are connected by the cylinder bordism $W = M \times [0, 1]$ , so any two oriented collapse maps for $M$ are homotopic. $□$

Proposition (Surjectivity of $Φ^{S O}$ via oriented transversality). Every class $[ϕ] \in π_{n + k} (M S O (k))$ is the oriented Pontryagin-Thom class of some closed smooth oriented $n$ -manifold $M$ .

Proof. Choose a representative $ϕ : S^{n + k} \to M S O (k) = Th (γ_{S O (k)}^{k})$ . By the transversality theorem (Thom's oriented version), perturb $ϕ$ to a smooth map transverse to the zero section $B S O (k) ↪ Th (γ_{S O (k)}^{k})$ . The preimage $$ M := \phi^{-1}(BSO(k)) \subset S^{n+k} $$ is a closed smooth submanifold of dimension $(n + k) - k = n$ , with normal bundle $ν_{M} = ϕ^{*} γ_{S O (k)}^{k}$ which is canonically oriented because $γ_{S O (k)}^{k}$ is. The orientation of $M$ comes from the orientation of $S^{n + k}$ together with the orientation of $ν_{M}$ . The Pontryagin-Thom construction applied to the oriented embedding $M ↪ S^{n + k}$ with this oriented normal bundle recovers the original class $[ϕ]$ . Different transverse perturbations yield oriented-bordant manifolds (parametrised oriented transversality applied to the cylinder $S^{n + k} \times [0, 1]$ ), so $Φ^{S O} ([M])$ is uniquely defined as $[ϕ]$ . $□$

Proposition (Computation of $Ω_{n}^{S O}$ for $n \leq 8$ ). With basis indexed by monomials in $[CP^{2}], [CP^{4}]$ of appropriate degree, plus the Wu-manifold $Z /2$ -torsion in dimension five (Wall 1960): $$ \Omega^{SO}_0 = \mathbb{Z}, \quad \Omega^{SO}_1 = \Omega^{SO}_2 = \Omega^{SO}_3 = 0, $$ $$ \Omega^{SO}_4 = \mathbb{Z} \langle [\mathbb{CP}^2] \rangle, \quad \Omega^{SO}_5 = \mathbb{Z}/2 \langle [W^5] \rangle, $$ $$ \Omega^{SO}_6 = \Omega^{SO}_7 = 0, \quad \Omega^{SO}_8 = \mathbb{Z}^2 \langle [\mathbb{CP}^4], [\mathbb{CP}^2 \times \mathbb{CP}^2] \rangle. $$

Proof. The rational computation $Ω_{*}^{S O} \otimes Q = Q [CP^{2}, CP^{4}, \dots]$ pins down the free rank: zero in dimensions not divisible by four; one in dimension four ( $CP^{2}$ ); zero in dimensions one through three, five through seven; two in dimension eight ( $CP^{4}$ and $CP^{2} \times CP^{2}$ , algebraically independent because their Pontryagin-number matrix $(p_{I} (CP^{4}); p_{(1, 1)} (CP^{4}))$ versus $(p_{I} (CP^{2})^{2}; (p_{I} (CP^{2}))^{2})$ is non-degenerate — explicitly $p_{2} (CP^{4}) = 10$ , $p_{1} (CP^{4})^{2} = 25$ , while $p_{1} (CP^{2} \times CP^{2})^{2} = 18$ and $p_{2} (CP^{2} \times CP^{2}) = 9$ via Künneth, giving distinct rational Pontryagin-character columns).

The torsion is computed by Wall 1960 ^[Wall] using a small-prime Adams spectral sequence: only the prime two contributes, and the spectral sequence is sparse below dimension eight, leaving exactly the $Z /2$ -class in $Ω_{5}^{S O}$ . The Wu-manifold $W^{5} = S U (3) / S O (3)$ is a five-dimensional homogeneous space whose total Stiefel-Whitney class restricted to its mod-2 cohomology gives $w (T W^{5}) = 1 + w_{2} + w_{3} + w_{2} w_{3}$ with $w_{2} \cdot w_{3} \neq = 0$ on the fundamental class; hence the Stiefel-Whitney number $w_{2} w_{3} [W^{5}] = 1 \in F_{2}$ is non-zero, so $[W^{5}] \neq = 0$ in $Ω_{5}^{S O}$ . By Wall's computation $Ω_{5}^{S O} ≅ Z /2$ generated by this manifold. $□$

Proposition (Hirzebruch signature identity at the generators). The signature homomorphism $σ : Ω_{4 k}^{S O} \otimes Q \to Q$ satisfies $σ ([CP^{2 k}]) = 1$ for every $k \geq 1$ .

Proof. The intersection form on $H^{2 k} (CP^{2 k}; Z) = Z ⟨ H^{k} ⟩$ is the rank-one matrix $⟨ 1 ⟩$ , where $H \in H^{2} (CP^{2 k})$ is the hyperplane class and $⟨ H^{2 k}, [CP^{2 k}]⟩ = 1$ . The signature of a rank-one positive-definite form is $1$ , so $σ (CP^{2 k}) = 1$ for every $k \geq 1$ . Combined with the polynomial-generator theorem, this means the signature is a ring homomorphism $Ω_{*}^{S O} \otimes Q \to Q$ sending every generator to $1$ , which is exactly the rational Hirzebruch $L$ -genus by uniqueness of multiplicative-sequence ring homomorphisms with $L ([CP^{2 k}]) = 1$ . $□$

Connections [Master]

Pontryagin and Chern classes 03.06.04. Provides the prerequisite framework for: Pontryagin classes of an oriented vector bundle, their relation to Chern classes via the complexification formula $p (E_{R}) \otimes C = c (E \oplus \overset{ˉ}{E})$ , and the Pontryagin-number pairing against the integral fundamental class. The Pontryagin-number bordism-invariance proof depends critically on the Whitney sum formula and stability under product line bundles.
Stiefel-Whitney and Pontryagin numbers 03.06.10. Sibling unit in the chapter. Both Stiefel-Whitney and Pontryagin numbers are introduced there; the present unit shows that Pontryagin numbers are the central invariants of oriented bordism, while Stiefel-Whitney numbers handle the 2-primary torsion. Together, 03.06.10 and the present unit identify the full set of bordism invariants for oriented manifolds.
Unoriented bordism and Thom's theorem 03.06.12. Direct sibling unit. Generalises in: replacing $B S O$ by $B O$ produces the unoriented Thom spectrum $M O$ , whose homotopy ring is the unoriented bordism ring $N_{*} = F_{2} [x_{n} : n + 1 \neq = 2^{k}]$ . The arguments mirror each other almost verbatim — oriented transversality and integral Stokes replace unoriented transversality and mod-2 Stokes — but the resulting rings are very different: $N_{*}$ is concentrated in characteristic two, while $Ω_{*}^{S O}$ has integral structure plus 2-torsion.
Hirzebruch signature theorem 03.06.11. The downstream specialisation in: once the rational structure of oriented bordism is known, the signature is exposed as a particular rational ring homomorphism $σ : Ω_{*}^{S O} \otimes Q \to Q$ computed via the $L$ -genus, which is the canonical multiplicative sequence in Pontryagin classes. The signature is even integer-valued on the integral bordism ring, not just rationally — this is one of the first instances of Lefschetz-style integrality on a manifold invariant.
Multiplicative sequences and the $L$ -, $A$ -, $Td$ -genera 03.06.15. The chapter-closing synthesis appears in: Hirzebruch's multiplicative-sequence formalism reinterprets ring homomorphisms out of $Ω_{*}^{S O}$ as formal power series in Pontryagin classes. The signature ( $L$ -genus), the $A$ -genus (spin index), and the Todd genus (complex case) are the canonical examples; each picks out a specific ring map detecting different bordism invariants.
Chern-Weil homomorphism 03.06.06. Provides the prerequisite framework for: characteristic-class representatives at the de Rham level, which combined with the present unit's bordism invariance produce all the closed-form integral formulae for cobordism characters (e.g., the integral of $L (p_{*})$ against the fundamental class is the signature in the parallel theory $σ : Ω_{*}^{S O} \to Z$ ).
Complex vector bundle 03.05.08. Builds toward: complex cobordism $Ω_{*}^{U}$ via $M U$ is the universal complex-oriented cohomology theory; Quillen 1969 identified its formal group law with the universal formal group law, making the present unit's $Ω_{*}^{S O}$ -computation a stepping stone toward a much richer additive structure visible in chromatic stable homotopy theory.
Whitney duality and immersion obstructions 03.06.16. Whitney duality on tangent and normal bundles produces dual families of Stiefel-Whitney numbers that are bordism invariants in $Ω_{*}^{O}$ , and the orientation-forgetting map $Ω_{*}^{S O} \to Ω_{*}^{O}$ from the present unit routes oriented bordism data into the immersion-obstruction theory of 03.06.16. Anchor phrase: Whitney duality routes through the orientation-forgetting map of the present unit.
Combinatorial Pontryagin classes and exotic 7-spheres 03.06.17. Milnor's exotic-sphere construction lives in the oriented category: $Σ^{7}$ is an oriented closed manifold, the bounding $8$ -manifold is oriented, and the signature formula on the boundary uses oriented bordism invariance of Pontryagin numbers — exactly the structural input proved here. The exotic-sphere group $Θ_{7}$ sits inside the long exact sequence of structure-group reductions over $Ω_{*}^{S O}$ . Anchor phrase: exotic spheres detect oriented-bordism invariants.
Signature of a $4 k$ -manifold and the intersection form 03.06.19. The signature is the canonical $Z$ -valued ring homomorphism $σ : Ω_{*}^{S O} \to Z$ , factoring through the rational-bordism structure computed here. The intersection-form treatment of 03.06.19 is the cohomological side of the bordism-invariance statement that the present unit proves. Anchor phrase: signature factors through the oriented-bordism ring of the present unit.
Borel-Hirzebruch and the cohomology of $G / T$ 03.06.20. The Thom spectrum $M S O$ has $H^{*} (M S O; Q) = H^{*} (B S O; Q) = Q [p_{1}, p_{2}, \dots]$ , identified by 03.06.20 as the Weyl-invariant subring of the polynomial ring on Pontryagin roots. The Borel-Hirzebruch computation supplies the cohomological data on which the present unit's Pontryagin-Thom argument lifts to the homotopy ring $π_{*} (M S O)$ .

Historical & philosophical context [Master]

Pontryagin 1947, 1950 ^[Rohlin] introduced the bordism relation in framed form to compute homotopy groups of spheres; the framed analogue $Ω_{n}^{fr}$ equals the stable homotopy group $π_{n}^{s}$ , and Pontryagin used this to compute $π_{n + 1} (S^{n})$ and $π_{n + 2} (S^{n})$ . Rokhlin 1952 ^[Rohlin] proved the signature divisibility theorem on spin four-manifolds — the precursor to the integral structure of $Ω_{4}^{S O}$ — using a clever argument with the intersection form modulo sixteen. Thom 1954 ^[Thom], in his Comment. Math. Helv. paper Quelques propriétés globales des variétés différentiables, abstracted the Pontryagin-Thom construction to arbitrary structure groups, introduced the Thom spectra $M O, M S O, M U$ , and gave the first complete computations: the unoriented case $N_{*}$ as a polynomial $F_{2}$ -algebra and the rational oriented case $Ω_{*}^{S O} \otimes Q$ as a polynomial $Q$ -algebra on ${[CP^{2 k}]}$ . Averbuch 1959 ^[Averbuch] computed $Ω_{4}^{S O} ≅ Z$ independently.

Wall 1960 ^[Wall], in Determination of the cobordism ring (Ann. Math. 72, 292-311), gave the integral computation of $Ω_{*}^{S O}$ through dimension eight, identifying all the 2-primary torsion via an explicit small-prime Adams spectral sequence and exhibiting the Wu manifold $W^{5} = S U (3) / S O (3)$ as the generator of $Ω_{5}^{S O} ≅ Z /2$ . Hirzebruch 1956 ^[Hirzebruch] Topological Methods in Algebraic Geometry (Springer-Verlag) bundled the rational Pontryagin-number computation into the signature theorem $σ (M^{4 k}) = ⟨ L_{k} (p_{1}, \dots, p_{k}), [M]⟩$ via the multiplicative-sequence $L$ -genus, providing the canonical reformulation. Milnor-Stasheff 1974 ^{[Milnor-Stasheff]} gave the modern textbook exposition.

The lineage continues in three directions. Novikov 1962 ^[Novikov] Mat. Sb. extended the Adams spectral sequence at all primes to compute torsion in $Ω_{*}^{S O}$ in higher dimensions, and also opened the complex-cobordism program leading to Quillen 1969's formal-group-law identification. Atiyah 1961 reformulated cobordism as a generalised cohomology theory $Ω_{*}^{S O} (X)$ on topological spaces, making the spectrum-level structure first-class. Wall 1966 and subsequent work extended the integral computation to all dimensions, identifying the higher torsion. The historical narrative — from Pontryagin's framed cobordism, through Thom's general structure-group abstraction, to Wall's integral computation and Hirzebruch's $L$ -genus reformulation — is the canonical model for every cobordism-theoretic invariant.

Bibliography [Master]

@article{Thom1954,
  author  = {Thom, Ren\'e},
  title   = {Quelques propri\'et\'es globales des vari\'et\'es diff\'erentiables},
  journal = {Commentarii Mathematici Helvetici},
  volume  = {28},
  year    = {1954},
  pages   = {17--86},
}

@article{Wall1960,
  author  = {Wall, C. T. C.},
  title   = {Determination of the cobordism ring},
  journal = {Annals of Mathematics},
  volume  = {72},
  year    = {1960},
  pages   = {292--311},
  note    = {Integral computation of \(\Omega^{SO}_*\) through dimension 8.},
}

@book{MilnorStasheff1974,
  author    = {Milnor, John W. and Stasheff, James D.},
  title     = {Characteristic Classes},
  publisher = {Princeton University Press},
  year      = {1974},
  series    = {Annals of Mathematics Studies},
  number    = {76},
  note      = {Pontryagin-Thom construction in \S17; computation of \(N_*\) in \S18.},
}

@book{Stong1968,
  author    = {Stong, Robert E.},
  title     = {Notes on Cobordism Theory},
  publisher = {Princeton University Press},
  year      = {1968},
  series    = {Mathematical Notes},
}

@book{Hirzebruch1956,
  author    = {Hirzebruch, Friedrich},
  title     = {Topological Methods in Algebraic Geometry},
  publisher = {Springer-Verlag},
  year      = {1956},
  note      = {\S1.5: Pontryagin numbers and the signature theorem on \(\Omega^{SO}_*\).},
}

@article{Novikov1962,
  author  = {Novikov, Sergei P.},
  title   = {Homotopy properties of Thom complexes},
  journal = {Matematicheskii Sbornik (N.S.)},
  volume  = {57(99)},
  year    = {1962},
  pages   = {407--442},
  note    = {Adams spectral sequence for \(MSO\); 2-primary torsion in higher-dimensional oriented bordism.},
}

@article{Rohlin1952,
  author  = {Rohlin, V. A.},
  title   = {A new result in the theory of 4-dimensional manifolds},
  journal = {Doklady Akademii Nauk SSSR},
  volume  = {84},
  year    = {1952},
  pages   = {221--224},
  note    = {Signature divisibility on spin 4-manifolds; precursor to integral \(\Omega^{SO}_4\).},
}

@article{Averbuch1959,
  author  = {Averbuch, B. G.},
  title   = {On the bordism ring of an orientable manifold},
  journal = {Doklady Akademii Nauk SSSR},
  volume  = {125},
  year    = {1959},
  pages   = {11--14},
}

@book{Milnor1965,
  author    = {Milnor, John W.},
  title     = {Topology from the Differentiable Viewpoint},
  publisher = {University Press of Virginia},
  year      = {1965},
  note      = {Pontryagin-Thom for framed cobordism in geometric form; oriented collaring theorem.},
}

Prerequisites

03.06.04
03.06.10
03.06.12

Tier anchors

beginner: Hatcher §3.D (bordism informal); Milnor-Stasheff §17 first half; pair-of-pants picture upgraded to oriented strip with consistent arrows
intermediate: Milnor-Stasheff *Characteristic Classes* §17 (Pontryagin-Thom construction in the oriented case); Stong *Notes on Cobordism Theory* Ch. V (oriented bordism); Hirzebruch *Topological Methods in Algebraic Geometry* §1.5
master: Thom 1954 *Comment. Math. Helv.* 28 *Quelques propriétés globales des variétés différentiables* (oriented case in the same paper as the unoriented); Wall 1960 *Ann. Math.* 72, 292-311 *Determination of the cobordism ring* (integral computation of Ω^SO_* up to dimension 8); Milnor-Stasheff §17 (Pontryagin-Thom collapse for MSO); Stong *Notes on Cobordism Theory* (Princeton, 1968) Ch. V; Novikov 1962 *Izv. Akad. Nauk SSSR* 26 (complex cobordism)

References

TODO_REF
Thom — Quelques propriétés globales des variétés différentiables · Comment. Math. Helv. 28 (1954), 17-86 (oriented bordism in §IV.13-15)
TODO_REF
Wall — Determination of the cobordism ring · Ann. Math. 72 (1960), 292-311 (integral computation of Ω^SO_* through dimension 8)
TODO_REF
Milnor-Stasheff — Characteristic Classes · Princeton University Press, 1974, §17 (Pontryagin-Thom) and §19 (oriented case calculations)
TODO_REF
Stong — Notes on Cobordism Theory · Princeton University Press, 1968, Ch. V (oriented cobordism)
TODO_REF
Hirzebruch — Topological Methods in Algebraic Geometry · Springer, 1956, §1.5 (signature theorem and L-genus on Ω^SO_*)
TODO_REF
Novikov — Homotopy properties of Thom complexes · Mat. Sb. (N.S.) 57(99) (1962), 407-442 (Adams spectral sequence for MSO; oriented cobordism at the prime 2)
TODO_REF
Averbuch — On the bordism ring of an orientable manifold · Dokl. Akad. Nauk SSSR 125 (1959), 11-14 (early computation of Ω^SO_4 ≅ ℤ)
TODO_REF
Rohlin — A new result in the theory of 4-dimensional manifolds · Dokl. Akad. Nauk SSSR 84 (1952), 221-224 (signature divisibility on spin 4-manifolds; precursor to integral structure of Ω^SO_4)

Lean module

Codex.Modern.CharClasses.OrientedBordismPT

Reviewer

TBD — characteristic-classes cluster reviewer; flag for oriented-bordism / Hirzebruch L-genus cross-check

Estimated time

beginner: 22m
intermediate: 48m
master: 95m

Intuition [Beginner]

Visual [Beginner]

Worked example [Beginner]

Check your understanding [Beginner]

Formal definition [Intermediate+]

Counterexamples to common slips [Intermediate+]

Key theorem with proof [Intermediate+]

Exercises [Intermediate+]

Lean formalization [Intermediate+]

Advanced results [Master]

The oriented bordism ring Ω∗SO​

Pontryagin-Thom via the MSO Thom spectrum

Thom's main theorem Ω∗SO​≅π∗​(MSO) and the rational computation

Wall 1960 — Pontryagin and Stiefel-Whitney numbers classify Ω∗SO​

Full proof set [Master]

Connections [Master]

Historical & philosophical context [Master]

Bibliography [Master]

The oriented bordism ring $Ω_{*}^{S O}$

Thom's main theorem $Ω_{}^{S O} ≅ π_{} (M S O)$ and the rational computation

Wall 1960 — Pontryagin and Stiefel-Whitney numbers classify $Ω_{*}^{S O}$