Alphabeta Math
TheoremStatement: AI-adaptedProof: AI-adaptedPipeline-generatedjudge pass (gpt-6.1-sol)audited 2026-10-08
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

The upper endpoint of a c-Cambrian fiber, interval fibers and the explicit formula u_c(w) = pi_{c^{-1}}(ww0)w0

Statement

Let (W,S) be a Coxeter system of finite type with longest element w0, let c be a Coxeter element, πc the sortable projection, ∼c the sortable equivalence and W/∼c the sortable quotient of The sortable projection kernel and the c-Cambrian quotient, and let ∧,∨ be the weak-order lattice operations on W (Weak order is a meet-semilattice, finite Coxeter groups are lattices, and joins of simple reflections exist exactly for finite parabolics). For J⊆S write wJ for the WJ-prefix and Jw0:=w0(J)w0 for the minimal representative of WJw0 (The weak parabolic projection, its adjoints, and the cover-join lemmas (2)). Define the upper projection of c by uc ⁣:W→W,uc(w):=πc−1(ww0) w0, where c−1 is the Coxeter element inverse to c, with the reversed reduced word, and πc−1 its sortable projection (The longest element as the opposition of the chamber, and longest elements of finite parabolics (1), The recursive initial-letter sortable projection). Whenever a recursion is indexed by a Coxeter element of a standard parabolic, its projections are formed there; in particular, usc in (1)(iii) is formed on WJ with longest element w0(J). Then:

(1) The terminal formula for πc and the recursions for uc. Let s∈S and J:=S∖{s}. (i) If s is final in c and ℓ(sw)<ℓ(w), then πc(w)=s∨πcs(wJ), where cs is the restriction of c to WJ obtained by deleting the final letter. (ii) If s is final in c and ℓ(sw)>ℓ(w), then uc(w)=s⋅uscs(sw). (iii) If s is initial in c and ℓ(sw)>ℓ(w), then uc(w)=sw0∧(usc(wJ)⋅Jw0).

(2) Monotonicity and idempotence of uc. The map uc is order preserving and idempotent, and w≤Ruc(w) for every w∈W.

(3) Fibers are closed intervals with these endpoints. For all x,y∈W, πc(x)=πc(y)  ⟺  uc(x)=uc(y),πc(uc(w))=πc(w),uc(πc(w))=uc(w). Consequently every ∼c-fiber is the closed interval [w]c={y∈W:πc(w)≤Ry≤Ruc(w)} with lower endpoint πc(w) and upper endpoint uc(w): no fiber has a gap, both endpoint maps are order preserving, and by the interval criterion The interval criterion for a lattice congruence: interval classes with monotone endpoints the equivalence ∼c is recovered from the two monotone endpoint maps as a lattice congruence — the same congruence of Sortable elements form a sublattice and the c-Cambrian quotient is its lattice-homomorphic image (4), now with its classes exhibited as the fibers.

(4) Abstention. The quotient is still not identified with the least lattice congruence contracting the oriented rank-two pairs of c, and no noncrossing, cluster-fan or counting statement is made. No Choice is used.

Facts & Assumptions

Given: a finite-type Coxeter system (W,S), its longest element w0, a Coxeter element c, its inverse c−1 represented by the reversed reduced word, the sortable projections πc and πc−1, the right weak order ≤R, and the parabolic prefixes and longest elements.

[F1]

The sortable projection kernel and the c-Cambrian quotient (1)-(2) and Finite lattice congruences, interval endpoints and descending rooted-chain labels (1): ∼c is the kernel relation x∼cy  ⟺  πc(x)=πc(y); W/∼c, pc, its quotient order and the proposed class meet/join operations are defined there.

[F2]

The recursive initial-letter sortable projection: if s is initial in c, then πc(w)=sπscs(sw) when ℓ(sw)<ℓ(w) and πc(w)=πsc(wJ) when ℓ(sw)>ℓ(w), with J=S∖{s} and wJ the WJ-prefix; the inverse Coxeter element uses the reversed word.

[F3]

c-sortable elements, forced and unforced skips, skip roots, and the chamber cone (1) and Coxeter elements, the oriented Euler form, the skew form, and the periodic word: a one-letter simple generator is c-sortable because its sorting word lies in the first block of c∞.

[F4]

The weak parabolic projection, its adjoints, and the cover-join lemmas (1)-(3): N(wJ−1)=N(w−1)∩ΦJ,+; wJ is the greatest WJ-element below w and prefix projection is order-preserving; the prefix projection preserves joins; and its largest lift is zw0(J)w0.

[F5]

The longest element as the opposition of the chamber, and longest elements of finite parabolics (1)(i)-(v): w02=1, ρ(w0)Φ+=Φ−, N(w0v)=Φ+∖N(v), ℓ(w0w)=ℓ(ww0)=ℓ(w0)−ℓ(w), and conjugation by w0 permutes S.

[F6]

The right and left weak orders, intervals, covers, and meets and joins of subsets (1),(3) and Weak order is a partial order with finite graded intervals; covers and the inversion-set criterion (2),(4)-(5): x≤Ry means y=xv with additive length; a simple left multiplication changes length by 1 or −1; s≤Rw iff ℓ(sw)<ℓ(w); and x≤Ry iff N(x−1)⊆N(y−1).

[F7]

The geometric inversion set N(w) of an element of a Coxeter group (1)-(2): N(w)={α∈Φ+:ρ(w)α∈Φ−}, with the positive and negative root partition and the inversion-set convention used in the proof.

[F9]

The recursive projection is well defined, sortable-valued, below w, idempotent, descent-detecting and parabolic (1)-(4): πc(w) is well-defined, c-sortable and below w; it fixes sortable elements and is idempotent; and for an initial letter s, w≥Rs iff πc(w)≥Rs.

[F10]

Omega-positive words are commutation-equivalent to sortable sorting words; sortable equals aligned; parabolic restriction (3): if v is c-sortable, its WJ-prefix is sortable for the restricted Coxeter element on WJ; conversely a sortable element of WJ is c-sortable in W.

[F11]

Skip bases, cover roots, greatest-sortable projections, and the chamber union of each cone (1): πc(w) is the unique greatest c-sortable element below w.

[F12]

Skip roots form a basis, negative skips are cover roots, and the cover decomposition of sortable elements (5)(ii): if s is final in c and v is c-sortable with v≥Rs, then v=s∨vJ.

[F13]

The cone criterion, monotonicity of the projection, and the greatest sortable element below w (2): πc is order-preserving; the same holds for any Coxeter element of a finite parabolic subsystem.

[F14]

Sortable elements form a sublattice and the c-Cambrian quotient is its lattice-homomorphic image (2): c-sortable elements are closed under nonempty joins, and their joins are c-sortable.

[F15]

Lattice quotient descent, class intervals and monotone endpoints (iii): for a finite lattice congruence, the proposed quotient operations are representative-independent and the quotient map preserves meet and join.

[F16]

The interval criterion for a lattice congruence: interval classes with monotone endpoints (i)-(ii): for an equivalence relation on a finite lattice whose classes are intervals, the relation is a congruence if and only if its lower and upper endpoint maps are order-preserving.

[F17]

The longest element as the opposition of the chamber, and longest elements of finite parabolics (2), applied to WJ: w0(J) is the longest element of the finite parabolic WJ and is an involution; applying the opposition assertion of [F5] within WJ gives N(w0(J)v)=ΦJ,+∖N(v) for v∈WJ.

[F18]

Sortable elements form a sublattice and the c-Cambrian quotient is its lattice-homomorphic image (4): the kernel relation ∼c is a lattice congruence with the quotient operations and quotient map already defined in The sortable projection kernel and the c-Cambrian quotient.

Proof

technique · derive the upper-projection recursions from the terminal projection formula and the longest-element anti-isomorphism; compare fibers by induction on rank and length; then identify each fiber as an interval and apply the finite-lattice interval criterion
1.1F4F5F6F7F17givenalgebra

Fix J=S∖{s} and let w=wJ⋅Jw be the length-additive parabolic factorization. The prefix inversion formula in [F4] and opposition in [F5] give N(((ww0)J)−1)=N((ww0)−1)∩ΦJ,+=N(w0w−1)∩ΦJ,+=ΦJ,+∖(N(w−1)∩ΦJ,+)=ΦJ,+∖N(wJ−1). Applying opposition inside WJ gives N(w0(J)wJ−1)=ΦJ,+∖N(wJ−1)=N((wJw0(J))−1). Both (ww0)J and wJw0(J) lie in WJ, so equality of their inverse inversion sets gives equality of the elements by the order criterion and antisymmetry in [F6]. Hence (ww0)J=wJw0(J).

1.2F3F4F10F11F12F14givenalgebra

Suppose s is final in c and ℓ(sw)<ℓ(w); set q:=πc(w) and v:=πcs(wJ). By [F3] and [F10], both s and v are c-sortable and lie below w, since s≤Rw and v≤RwJ≤Rw. Their join x:=s∨v is c-sortable by [F14] and below w, so x≤Rq by [F11]. Thus q≥Rs; the terminal cover decomposition [F12] gives q=s∨qJ. The prefix qJ is cs-sortable by [F10] and qJ≤RwJ by [F4], hence qJ≤Rv by [F11] applied inside WJ. Therefore q=s∨qJ≤Rs∨v=x, and with x≤Rq this proves πc(w)=s∨πcs(wJ).

1.3baseih

We prove by lexicographic induction on (∣S∣,ℓ(x)) that whenever x≤Ry and πc(x)=πc(y), one has uc(x)=uc(y). If S=∅, then W={1} and this holds; at every positive-rank pair assume it holds for all smaller measures, and fix an initial letter s of c.

1.4F5F6F9F13givenalgebra

Define τ(w):=ww0, so uc=τ∘πc−1∘τ. The map τ reverses right weak order: if y=xv with additive length, then xw0=yw0 (w0v−1w0) and ℓ(w0v−1w0)=ℓ(v) by [F5], so yw0≤Rxw0 by [F6]; since τ2 is the identity, this is an order anti-isomorphism. Therefore uc is order-preserving by [F13]. Since πc−1(ww0)≤Rww0 by [F9], applying τ gives w≤Ruc(w). Finally, uc(uc(w))=τ(πc−1(πc−1(ww0)))=τ(πc−1(ww0))=uc(w) by idempotence in [F9].

2.1F2F5F6F8F17step 1.1step 1.2step 1.4givenalgebra

If s is final in c and ℓ(sw)>ℓ(w), then s is initial in c−1 and ℓ(sww0)<ℓ(ww0) by [F5]. The initial-letter recursion [F2] gives πc−1(ww0)=sπ(scs)−1((sw)w0), so uc(w)=s uscs(sw), proving (1)(ii). If s is initial in c and ℓ(sw)>ℓ(w), then s is final in c−1 and ℓ(sww0)<ℓ(ww0); apply step 1.2 to c−1 and ww0 to get πc−1(ww0)=s∨π(sc)−1((ww0)J). Multiplying on the right by w0 converts the join to a meet by the order reversal in step 1.4; using (ww0)J=wJw0(J) from step 1.1 and w0(J)2=1 gives uc(w)=sw0∧(π(sc)−1(wJw0(J))w0)=sw0∧(usc(wJ) Jw0), where usc is formed inside WJ with longest element w0(J) and Jw0=w0(J)w0. This proves (1)(iii).

3.1F2F4F6F9step 1.3step 2.1givenalgebra

Continue the induction of step 1.3. Suppose x≤Ry and πc(x)=πc(y). If ℓ(sx)<ℓ(x), then ℓ(sy)<ℓ(y) because s≤Rx≤Ry by [F6]. Write y=xv with additive length. Then sy=(sx)v and ℓ(sy)=ℓ(sx)+ℓ(v), so sx≤Rsy. The recursion [F2] gives πscs(sx)=πscs(sy); since ℓ(sx)=ℓ(x)−1, the induction hypothesis yields uscs(sx)=uscs(sy). In scs, the letter s is final and sx,sy have left ascent s, so step 2.1(ii) gives uscs(sx)=suc(x) and uscs(sy)=suc(y); cancellation proves uc(x)=uc(y). If instead ℓ(sx)>ℓ(x) but ℓ(sy)<ℓ(y), then πc(x)=πsc(xJ)≤Rx lies outside the filter above s, while πc(y)=sπscs(sy) lies in that filter: indeed sy̸≥Rs, and πscs(sy)≤Rsy by [F9], so left multiplication by s lengthens this projection by one. This contradicts πc(x)=πc(y). Thus both are left ascents. The parabolic prefix is order-preserving by [F4], so xJ≤RyJ; the recursion gives πsc(xJ)=πsc(yJ), and the induction hypothesis in lower rank gives usc(xJ)=usc(yJ). Formula 2.1(iii), with the same sw0 and Jw0 for both inputs, now yields uc(x)=uc(y). This completes the comparable-pair induction.

4.1F5F9step 3.1step 1.4givenalgebra

For arbitrary x,y with πc(x)=πc(y)=v, [F9] gives v≤Rx,y and πc(v)=v. Applying the comparable-pair result of step 3.1 to (v,x) and (v,y) gives uc(x)=uc(v)=uc(y). Conversely, if uc(x)=uc(y), then πc−1(xw0)=πc−1(yw0) by the definition of uc; the forward implication just proved for arbitrary pairs, applied to c−1, gives uc−1(xw0)=uc−1(yw0). By definition these are πc(x)w0 and πc(y)w0, so πc(x)=πc(y). Thus the two fiber partitions agree. The forward implication and idempotence of πc also give uc(πc(w))=uc(w). Applying this identity to c−1 and using uc(w)w0=πc−1(ww0) yields πc(uc(w))w0=uc−1(uc(w)w0)=uc−1(πc−1(ww0))=uc−1(ww0)=πc(w)w0, so πc(uc(w))=πc(w).

5.1F9F13step 1.4step 4.1givenalgebra

If y∈[x]c, then πc(y)=πc(x), so step 4.1 gives uc(y)=uc(x); by [F9] and step 1.4, πc(x)=πc(y)≤Ry≤Ruc(y)=uc(x). Conversely, if πc(x)≤Ry≤Ruc(x), monotonicity [F13] and step 4.1 give πc(x)=πc(πc(x))≤Rπc(y)≤Rπc(uc(x))=πc(x), so πc(y)=πc(x). Thus [x]c=[πc(x),uc(x)] with the asserted endpoints; the endpoint maps are order-preserving by [F13] and step 1.4.

6.1F1F15F16F18step 5.1discharge-inductiongivenalgebra∎

The classes are intervals by step 5.1, and their endpoint maps are order-preserving there; applying [F16] shows that ∼c is a lattice congruence. It is the same kernel relation and quotient as in [F1] and the congruence conclusion of [F18], not an identification with a different least-contraction congruence. The finite quotient consequences of [F15] give the representative-independent class operations and the lattice-homomorphic quotient map. All inductions are finite and no Choice is used.

Depends on

Used by

Cited to discharge well-definedness by The sortable projection kernel and the c-Cambrian quotient.

Dependency tree · two levels

57 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.

Sources