Theorem The outgoing-kernel character of a reduced word
✓ theorem proved

The outgoing-kernel character of a reduced word

Fix a finite-generator Coxeter system, a finite real basis indexed by \(\sigma\) of its extended realization, and a reduced word \(\mathbf t\) with product \(b\). Let \(P=\mathbb R[X_i:i\in\sigma]\) be the polynomial coefficient algebra for this basis, and let \(S_{\mathbf t}\) be its word sheaf on \([1,b]\). For \(x\le b\), put \[ K_x(S_{\mathbf t}) =\bigcap_{e:x\to y}\ker\rho^-_e,\qquad d(x,b)=\ell(b)-\ell(x). \] This is the full outgoing kernel.

It is finite free over \(P\). Its character \(K_x^{\mathbf t}(q)\) and the stalk character satisfy \[ K_x^{\mathbf t}(q) =\operatorname{reflect}_{d(x,b)}C_x^{\mathbf t}(q), \qquad \operatorname{natDegree}C_x^{\mathbf t}\le d(x,b). \] Thus the kernel character also has the reciprocal expression \(q^{d(x,b)}C_x^{\mathbf t}(q^{-1})\) in the Laurent polynomial ring.

Relation to the OpenAI notes

Different intermediate objects.

OpenAI notes: The normalization discussion identifies indecomposable boundary-sheaf characters with KL polynomials by the cited character theorem. bg:normalization.

Lean (primary): The word-stalk character is proved to equal a positive Hecke-word coefficient. Its R-convolution is then proved before model decomposition and cancellation. The separate boundary-character identification is exposed as StalkCharacterObligation. IntegralPairings:289, CharacterReciprocity:1134, SheafMorphisms:219, BottSamelson:1924, BottSamelson:1945, IntegralPairings:104, IntegralPairings:120, CharacterReciprocity:915.

Role in the proof: The word formula and R-convolution describe the tensor word. A common decomposition into boundary models then supplies the character identities used in the cancellation argument.

Proof

Step 1. Specify the free-sheaf and duality data

The source packages the polynomial word sheaf as bottFreeSheaf. Its fields are the sheaf itself, free stalks, the root-labelled upper-quotient property, and global generation.

The reduced-word section pairing supplies a perfect pairing with shift \(m=|\mathbf t|\) and both support-orthogonality equations. The cut-balance construction turns it into bottFreeDuality, the section-duality data for this free sheaf.

Step 2. Identify the full kernel with the stalk dual

The stalk–costalk theorem gives a \(P\)-linear equivalence \[ \Phi_x:(S_{\mathbf t}^x)^*\xrightarrow{\sim}K_x(S_{\mathbf t}), \] where the star denotes \(\operatorname{Hom}_P(-,P)\). The full kernel can be written either as bundled kernelData or as the kernel of the map taking all outgoing restrictions; their equivalence keeps the underlying vector.

Since the stalk is finite free, its dual is finite free. Transport through \(\Phi_x\) proves kernel freeness; finite generation also follows from this finite basis.

Step 3. Calculate the duality degree

There are \(\ell(x)\) incoming reflection edges at \(x\) in \([1,b]\). The inverse stalk–costalk degree theorem sends a homogeneous functional of degree \(r\) to kernel degree \[ r+m-\ell(x). \] Reducedness gives \(m=\ell(b)\), so this degree is \(r+d(x,b)\).

Step 4. Reflect the character and bound its degree

Apply the homogeneous dual-character theorem with \(M=S_{\mathbf t}^x\), \(N'=K_x(S_{\mathbf t})\), and \(D=d(x,b)\). The source kernel grading is nonnegative, and Step 3 verifies the homogeneity hypothesis.

For stalk basis degrees \(d_i\), the transported dual basis has degrees \(D-d_i\ge0\). Hence \(d_i\le D\), the stalk character has degree at most \(D\), and \[ K_x^{\mathbf t}(q)=\sum_iq^{D-d_i} =\operatorname{reflect}_D C_x^{\mathbf t}(q). \] The polynomial word-sheaf stalk character is the defining bottStalkCharacter, so this is the source's bottKernelCharacter identity.

Step 5. Read the reciprocal form

The proved degree bound allows the bounded reciprocal formula for polynomial reflection: \[ \operatorname{reflect}_D C_x^{\mathbf t}(q) =q^D C_x^{\mathbf t}(q^{-1}). \] Each exponent \(D-d_i\) is nonnegative, so this Laurent expression represents the ordinary kernel character polynomial.

References

The local source declarations below give the construction and proof.

Lean formalization

OAI.KLInvariance.TitsSpace.bottFreeSheaf: CharacterReciprocity.lean:897.

OAI.KLInvariance.TitsSpace.bottFreeDuality: CharacterReciprocity.lean:906.

OAI.KLInvariance.TitsSpace.bottKernelFree: CharacterReciprocity.lean:910.

OAI.KLInvariance.TitsSpace.bottKernelCharacter: CharacterReciprocity.lean:915.

OAI.KLInvariance.TitsSpace.bottStalkCharacter_natDegree_le: CharacterReciprocity.lean:1088.

Used by