---
title: Inversions Tableaux and Schubert Polynomials
url: https://www.emergentmind.com/topics/inversions-tableaux
type: topic
---

# Inversions Tableaux and Schubert Polynomials

Searching arXiv for the exact topic and closely related papers to ground the article in current literature.
{"query":"ti:\"Inversions Tableaux\" OR abs:\"Inversions Tableaux\"","max_results":5,"sort_by":"submittedDate","sort_order":"descending"}
{"query":"all:\"inversions tableaux\"","max_results":10,"sort_by":"submittedDate","sort_order":"descending"}
{"query":"all:\"tableau inversions\" OR all:\"inverted semistandard young tableaux\" OR all:\"combinatorics of tableau inversions\"","max_results":10,"sort_by":"submittedDate","sort_order":"descending"}
Inversions tableaux are a combinatorial model for Schubert polynomials and Stanley symmetric functions in which the underlying shape is the inversion diagram of a permutation and the filling is constrained by a weak balancing condition, column distinctness, and row or diagonal upper bounds. Introduced as a modification of Edelman–Greene balanced staircase tableaux, they directly specialize to semistandard Young tableaux in the Grassmannian case and support explicit descriptions of extremal monomials, dominant permutations, and certain skew constructions [2507.11516].

## 1. Definition on inversion diagrams

Fix $n \ge 2$. The ambient staircase array is indexed by matrix coordinates $(i,j)$ with $1 \le i < j \le n$, with rows increasing bottom to top and columns increasing left to right. For a permutation $w \in S_n$, its inversion set is
$$
\operatorname{Inv}(w)=\{(i,j)\mid i<j \text{ and } w_j<w_i\}.
$$
The inversions diagram $T_w$ is obtained by shading precisely those boxes $(i,j)$ that belong to $\operatorname{Inv}(w)$ [2507.11516].

An inversions tableau of shape $w$ is a filling of the shaded boxes of $T_w$ by entries in $\{1,2,\dots,n\}$, with unshaded boxes assigned the conventionally fixed value $0$, subject to three local conditions. The first is weak balance: for all $1 \le i < j < k \le n$, the entries satisfy
$$
T(i,j)\le T(i,k)\le T(j,k)\quad\text{or}\quad T(j,k)\le T(i,k)\le T(i,j).
$$
This is the Rectangle Rule. The second is column strictness: within a fixed column $j$, the entries $T(i,j)$ in shaded boxes are pairwise distinct. The third is the diagonal bound $T(i,i+1)\le i$. An equivalent form replaces the diagonal bound by the row bound that all entries in row $i$ are $\le i$; with weak balance and column strictness, these two formulations are equivalent [2507.11516].

The local axioms may be summarized as follows.

| Condition | Content |
|---|---|
| (IT1) | Weak balance via the Rectangle Rule |
| (IT2) | No repeated entries within a column |
| (IT3) | $T(i,i+1)\le i$ |
| (IT3′) | Equivalent form: all entries in row $i$ are $\le i$ |

The weight of an inversions tableau $T$ is
$$
\operatorname{wt}(T)=(m_1,m_2,\dots,m_{n-1}),
$$
where $m_r$ is the number of shaded boxes filled with $r$; unshaded boxes do not contribute. This weight is the exponent vector appearing in the Schubert polynomial expansion [2507.11516].

The model is explicitly positioned as a weakening and restriction of Edelman–Greene balanced staircase tableaux. Balanced staircase tableaux fill the full staircase with distinct entries satisfying a strict median condition, equivalently the Strict Rectangle Rule; inversions tableaux instead fill only the inversion diagram, allow repeated entries, impose the weak Rectangle Rule, and add column distinctness together with row or diagonal bounds [2507.11516].

## 2. Realization of Schubert polynomials and Stanley symmetric functions

The central theorem identifies the Schubert polynomial $\mathfrak{S}_w$ with the weight generating function of inversions tableaux:
$$
\mathfrak{S}_w(\mathbf{x})=\sum_{T\in IT(w)} x_1^{m_1(T)}x_2^{m_2(T)}\cdots x_{n-1}^{m_{n-1}(T)}.
$$
Here $IT(w)$ denotes the set of inversions tableaux of shape $w$ [2507.11516].

The proof proceeds through reduced pipe dreams. If $RP(w)$ denotes the set of reduced pipe dreams for $w$, Bergeron–Billey’s formula gives
$$
\mathfrak{S}_w(\mathbf{x})=\sum_{P\in RP(w)} x_1^{d_1(P)}x_2^{d_2(P)}\cdots x_n^{d_n(P)},
$$
where $d_i(P)$ is the number of crosses in row $i$. The bijection $\phi:RP(w)\to IT(w)$ labels each shaded box $(i,j)\in\operatorname{Inv}(w)$ by the row in which pipes $i$ and $j$ cross, and places $0$ in unshaded boxes. Under this correspondence, weak balance reflects the local compatibility required by reducedness, column distinctness comes from the impossibility of a pipe meeting two lower-numbered pipes in the same row, and the row bound comes from the fact that pipe $i$ cannot cross below row $i$ [2507.11516].

The inverse map reconstructs a reduced pipe dream from a tableau by refining the weak order of boxes into a balanced total order and inserting crosses in reverse order. Conditions (IT2) and (IT3′) guarantee uniqueness of the row placements. As a result, the tableau model is weight-preserving and bijective, rather than merely enumerative [2507.11516].

Stabilization yields a tableau model for Stanley symmetric functions. If $UIT(w)$ denotes the set of unbounded inversions tableaux, meaning that (IT1) and (IT2) are retained but (IT3) is omitted, then
$$
F_w(\mathbf{x})=\sum_{T\in UIT(w)} x^{\operatorname{wt}(T)}.
$$
This is the stable limit corresponding to $F_w(\mathbf{x})=\lim_{m\to\infty}\mathfrak{S}_{1^m\times w}(\mathbf{x})$ [2507.11516].

## 3. Grassmannian specialization and semistandard tableaux

For a $k$-Grassmannian permutation $w$, the inversion diagram becomes a Ferrers diagram. More precisely, if $w$ has unique descent at position $k$, then the shaded region of $T_w$ is exactly the Ferrers diagram of a partition $\lambda_w$ placed with top-left corner at $(k,k+1)$. In this case, inversions tableaux specialize directly to reverse semistandard Young tableaux: entries lie in $\{1,2,\dots,k\}$, rows weakly decrease from left to right, and columns strictly decrease from top to bottom [2507.11516].

Consequently, $IT(w)$ is in bijection with reverse semistandard Young tableaux of shape $\lambda_w$ with entries at most $k$, and one recovers the Schur specialization
$$
\mathfrak{S}_w(x_1,\dots,x_k)=s_{\lambda_w}(x_1,\dots,x_k).
$$
This is the direct specialization to semistandard Young tableaux emphasized in the definition of the model [2507.11516].

The inverse Grassmannian case is also described explicitly. If $w$ is $k$-Grassmannian, then $T_{w^{-1}}$ occupies the intersections of columns $a_1,\dots,a_k$ and rows $b_1,\dots,b_{n-k}$. After deleting empty rows and columns and rotating by $180^\circ$, $IT(w^{-1})$ corresponds to flagged semistandard Young tableaux of shape $\lambda'$ with row bounds $(b_1,\dots,b_{n-k})$, where
$$
\lambda'_i=k+i-b_i.
$$
This provides a second semistandard interpretation, now in a flagged setting [2507.11516].

A representative example is given by the permutation $w=3469\,12578$, which is $4$-Grassmannian and has $\lambda_w=(5,3,2,2)$. In that case, inversions tableaux are reverse semistandard Young tableaux of shape $(5,3,2,2)$ with entries in $\{1,2,3,4\}$, and their weight generating function is $s_{(5,3,2,2)}(x_1,x_2,x_3,x_4)$ [2507.11516].

## 4. Extremal monomials and dominant permutations

The model gives direct tableau descriptions of the lexicographically maximal and minimal monomials in a Schubert polynomial. If $\operatorname{code}(w)=(c_1,\dots,c_{n-1})$ is the Lehmer code, with $c_i$ equal to the number of shaded boxes in row $i$ of $T_w$, then the lexicographically maximal monomial of $\mathfrak{S}_w$ is
$$
x^{\operatorname{code}(w)},
$$
with coefficient $1$. The unique tableau realizing it fills every shaded box in row $i$ by the entry $i$ [2507.11516].

The lexicographically minimal monomial is described in terms of the column-Lehmer code $ccode(w)=(c_1,\dots,c_n)$, where $c_j$ is the number of shaded boxes in column $j$. Setting
$$
e_i=\#\{j: c_j\ge i\},
$$
the minimal monomial is
$$
x_1^{e_1}x_2^{e_2}\cdots x_{n-1}^{e_{n-1}},
$$
again with coefficient $1$. The corresponding tableau is obtained by filling each column bottom-up with the smallest possible entries that do not repeat within that column. The construction is columnwise greedy and is compatible with the Rectangle Rule and column distinctness [2507.11516].

The model also yields a sharp characterization of dominant permutations. A permutation is dominant if its Lehmer code is weakly decreasing, equivalently if it avoids the pattern $132$. In diagrammatic terms, this means that the inversion diagram is downward closed. For such $w$, there is a unique inversions tableau, namely the row-wise fill
$$
T(i,j)=i
$$
on every shaded box, and hence
$$
\mathfrak{S}_w=x^{\operatorname{code}(w)}.
$$
This is the tableau-theoretic form of the fact that dominant Schubert polynomials are single monomials [2507.11516].

For the concrete permutation $w=431562$, the paper exhibits three inversions tableaux, giving the three monomials
$$
x_1^3x_2^2x_4x_5,\qquad
x_1^3x_2^2x_3x_5,\qquad
x_1^3x_2^2x_3x_4.
$$
The first is the lexicographically maximal monomial and the third is the lexicographically minimal monomial [2507.11516].

## 5. Skew variants and algorithmic structure

The paper introduces tableaux skew Schubert polynomials in the Grassmannian setting. If $u$ and $w$ are $k$-Grassmannian with $u\le w$ in weak Bruhat order, equivalently $\lambda_u\subseteq\lambda_w$, then one defines a skew inversions diagram by taking the shaded boxes in $\operatorname{Inv}(w)\setminus\operatorname{Inv}(u)$. Boxes of $\operatorname{Inv}(u)$ are filled with $k$, unshaded boxes are filled with $0$, and the remaining boxes are filled with entries in $\{1,\dots,k\}$ satisfying the same local conditions (IT1)–(IT3) [2507.11516].

If $IT(w/u)$ denotes this set, then the tableaux skew Schubert polynomial is
$$
\mathfrak{S}^{t}_{w/u}(\mathbf{x})=\sum_{T\in IT(w/u)} x^{\operatorname{wt}(T)}.
$$
In the Grassmannian case this coincides with the skew Schur polynomial:
$$
\mathfrak{S}^{t}_{w/u}(x_1,\dots,x_k)=s_{\lambda_w/\lambda_u}(x_1,\dots,x_k).
$$
Accordingly, it admits the Littlewood–Richardson expansion
$$
s_{\lambda_w/\lambda_u}=\sum_\nu c^{\lambda_w}_{\lambda_u,\nu}s_\nu.
$$
A plausible implication is that the skew construction is designed to extend Schubert-theoretic positivity phenomena beyond the ordinary Grassmannian case, although the non-Grassmannian extension is explicitly left open [2507.11516].

The model also admits two algorithmic descriptions. Algorithm A generates all inversions tableaux by enumerating reduced pipe dreams for $w$ and applying the bijection $\phi$. Algorithm B performs a direct tableau search column by column: for each column $j$, one assigns distinct entries $e_t\in\{1,\dots,i_t\}$ to the shaded boxes $(i_t,j)$ and checks the Rectangle Rule locally after each assignment. Since the entry domains are finite and column distinctness bounds the branching factor, the search terminates [2507.11516].

These constructions are tied to structural moves on pipe dreams. The paper records that generalized chute moves correspond to operations on inversions tableaux, and that columnwise monotonicity of associated Lehmer coordinates will be used in forthcoming work on semidistributive polygonal lattice structure for chute move posets. This suggests a structural role for inversions tableaux beyond weight enumeration [2507.11516].

## 6. Relation to other inversion-based tableau theories

The term “inversions tableaux” can be misleading because several earlier theories also combine tableaux with inversion statistics, but they concern different objects. In “Combinatorics of Tableau Inversions” [1412.6473], a tableau inversion is a pair of entries in the same column of a row-standard Young tableau that fail the order needed for column-standardness, and $i$-inverted tableaux are graded by the number of such pairs. “Inversions of Semistandard Young Tableaux” extends that framework to repeated entries and proves, among other results, a maximum inversion formula and invariance of $|S_i(\lambda,\mu)|$ under permutation of content [1508.01175]. “Generating Functions for Inverted Semistandard Young Tableaux and Generalized Ballot Numbers” develops per-standardization generating functions and Dyck-path interpretations for these inversion counts [1606.04869]. These theories study inversion statistics on Young tableaux; they do not define the inversion-diagram fillings that realize Schubert polynomials [1412.6473, 1508.01175, 1606.04869].

A second nearby but distinct line of work concerns permutation tableaux. “Permutation Tableaux and the Dashed Permutation Pattern 32–1” defines the inversion number of a permutation tableau via alternating paths and proves that it equals the number of occurrences of the dashed pattern $32\text{–}1$ in the reverse complement of the associated permutation [1007.5019]. In type $B$, “Permutation statistics and weak Bruhat order in permutation tableaux of type $B$” identifies Coxeter length with a tableau statistic involving typed zeros and $1$’s [1412.6290]. Here again, the objects are permutation tableaux rather than inversion-diagram tableaux [1007.5019, 1412.6290].

Within Schubert combinatorics, inversions tableaux are therefore best understood as a new tableau model obtained by modifying Edelman–Greene balanced staircase tableaux and restricting them to inversion diagrams, with the distinctive feature that their weight generating function is exactly $\mathfrak{S}_w$ and their stable version gives $F_w$ [2507.11516]. This places them at the intersection of pipe-dream combinatorics, Schur specialization in the Grassmannian regime, and local tableau rules modeled on weak balancing.

Source: https://www.emergentmind.com/topics/inversions-tableaux