Total Positivity and Symmetric Spaces
Huanchen Bao · 2026
Read Total Positivity and Symmetric Spaces by Huanchen Bao (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-25516.txt.
Introduction
Total positivity began with matrices whose every minor is nonnegative. Lusztig widened it into a rich theory for groups with a cell structure underneath. Huanchen Bao carries the theory onto symmetric spaces in 2026, the quotients formed when a group is folded by an involution. The totally nonnegative part of such a space is the closure of the image of the group's positive part. Double Bruhat cells, the intersections organising the group case, get space analogues with positive pieces.
Bao proves a full cell decomposition of the nonnegative symmetric space. Each cell gets an explicit positive parametrisation. Closure relations then say which cells touch which. Switching between the two natural coordinate families never needs subtraction. That trait is prized in the classical theory. A few symmetric pairs are excluded by a compatibility condition on the involution.
Catalog record
| CREATOR | Huanchen Bao |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.25516
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Total Positivity and Symmetric Spaces Huanchen Bao Abstract . We define a notion of total positivity for the symmetric space G / K G/K by taking the Hausdorff closure of the image of Lusztig’s totally positive part G > 0 G_{>0} in G / K G/K . We introduce double Bruhat cells for the symmetric space and define their totally positive pieces. We prove a cell decomposition of the totally nonnegative symmetric space, give explicit positive parametrizations of all cells, establish closure relations, and show that the transition maps between the two natural families of parametrizations are subtraction-free. 1. Introduction 1.1. Total positivity began with the classical study of matrices whose minors are all nonnegative or positive. Lusztig [ LuTotalPositivity ] extended this theory from matrices to arbitrary split connected reductive groups, using deep positivity properties coming from his theory of canonical bases. One of the key features of Lusztig’s theory is that total positivity is not merely a subset of real points; it comes with a rich cell structure. For G itself, the relevant cells are closely related to double Bruhat cells. Lusztig also defined totally nonnegative flag varieties and partial flag varieties. Marsh and Rietsch [ MarshRietschParametrizationsFlagVarieties ] gave explicit positive parametrizations of the cells in the full flag variety using positive subexpressions and generalized Chamber Ansatz coordinates, while Rietsch [ RietschClosureRelationsPartialFlag ] proved closure relations for totally nonnegative cells in partial flag varieties. The resulting theory of total positivity has become a central object connecting algebraic groups, canonical bases, cluster structures, Poisson geometry, and combinatorics. 1.2. The purpose of this paper is to develop the theory of total positivity for symmetric spaces. Let G be a split connected reductive group over \mathbb{R} , let \theta be an involution of G , and let K=G^{\theta} be the fixed-point subgroup. We study the symmetric space \mathcal{S}=G/K. In the Lusztig theory of total positivity, one needs to fix a pinning for the reductive group G , including Borel subgroups B^{\pm} . We further require (B^{+},T) to be a \theta -split pair in the sense of Springer [ SpringerAlgebraicGroupsWithInvolutions ] . In particular, \theta(B^{+})=\dot{w}^{\bullet}B^{+}\dot{w}^{\bullet,-1} for some Weyl group element w^{\bullet} . The involution \theta is further normalized using the \imath -root datum attached to the symmetric pair; the precise convention is recalled in § 2.1 . One should take this as the pinning data for the symmetric space. Throughout this paper we restrict to the class of symmetric pairs for which this normalized involution is compatible with the diagram involution, equivalently the parameters in § 2.1 may be chosen so that \overline{\zeta}_{i}=-1 . This condition excludes a few cases (see § 2.1 for the detailed list), but it is the natural setting in which the involution is compatible with the positivity structure used below. 1.3. The central geometric idea of the paper is to introduce the analogue of double Bruhat cells for the symmetric space. We use the embedding \iota:G/K\longrightarrow G,\qquad gK\longmapsto g\theta(g)^{-1}, which realizes G/K as a closed subvariety of G . Then, for u\leq w^{\bullet} in the Weyl group of G , we define \mathring{\mathcal{S}}_{u}=\iota^{-1}(B^{+}\dot{u}B^{-}\dot{w}^{\bullet})=\iota^{-1}(B^{+}\dot{u}\dot{w}^{\bullet}\theta(B^{-})). We call the strata \mathring{\mathcal{S}}_{u} the double Bruhat cells of the symmetric space. Note that \mathring{\mathcal{S}}_{u}=\emptyset if u\not\leq w^{\bullet} . This terminology is not only formal. These strata are motivated by Poisson geometry studied by Evens and Lu [ EvensLu2006 ] : they are the T -leaves for the natural Poisson structure on the symmetric space. Thus they play for G/K the role played by double Bruhat cells in G and by projected Richardson strata in partial flag varieties. In the special case where the symmetric space is the group itself, this construction recovers the usual double Bruhat cells in G . This Poisson-geometric interpretation is one of the main reasons these strata are the correct pieces on which to study total positivity. The relevant Frobenius splitting and cluster structure on the double Bruhat cell shall appear in forthcoming works. We now define the totally nonnegative symmetric space. Let \mathcal{S}_{\geq 0}=(G/K)_{\geq 0}:=\overline{G_{>0}K/K}, where the closure is taken in the Hausdorff topology on the real locus of G/K . For each u\leq w^{\bullet} , define \mathcal{S}_{u}^{>0}:=\mathcal{S}_{\geq 0}\cap\mathring{\mathcal{S}}_{u},\qquad\mathcal{S}_{u}^{\geq 0}:=\overline{\mathcal{S}_{u}^{>0}}. We also define the totally positive symmetric space by \mathcal{S}_{>0}:=\mathcal{S}_{e}^{>0}. The indexing set for these cells can be described in two equivalent ways. From the symmetric space side, the strata are indexed by Q=\{u\in W\mid u\leq w^{\bullet}\}. The second indexing set is motivated by partial flag varieties and the projected Richardson strata that appear there: P=\{(v,w)\mid v\in W_{I_{\bullet}},\ w\in W^{I_{\bullet}},\ v\leq w\}. The two descriptions are related by the map \phi:P\longrightarrow Q^{\mathrm{opp}},\qquad(v,w)\longmapsto v\theta(w^{-1})w^{\bullet}. The main combinatorial result of the paper is that this map is an isomorphism of posets; see Proposition 3.2 . Thus the symmetric space stratification is controlled by the same combinatorics that governs projected Richardson strata. The poset Q has a natural involution via the inverse map. We denote the induced involution on P by \sigma . 1.4. We can now state the main results. Let (v,w)\in P , choose a reduced expression \mathbf{w} of w , and let \mathbf{v}_{+} be the positive subexpression of v in \mathbf{w} in terms of Marsh and Rietsch [ MarshRietschParametrizationsFlagVarieties ] . Let G^{>0}_{\mathbf{v}_{+},\mathbf{w}} denote the corresponding Marsh–Rietsch positive parameter set. We further define the positive set H^{>0}_{\mathbf{v}_{+},\mathbf{w}} as the positive-root analogue of the Marsh–Rietsch set G^{>0}_{\mathbf{v}_{+},\mathbf{w}} ; it is defined precisely in §2.4 by applying the Chevalley involution to the Marsh–Rietsch parameter set. Let \overline{T}=T/T^{\theta} be the quotient torus and let \overline{T}_{>0} be its totally positive part. Theorem 1 (Theorem 3.5 ) . (1) The space \mathcal{S}_{\geq 0} is entirely contained in both the open B^{+} -orbit and the open B^{-} -orbit of G/K , that is, \mathcal{S}_{\geq 0}\subset(B^{-}K/K)\cap(B^{+}K/K). (2) For every (v,w)\in P , there is an isomorphism of semi-algebraic varieties G^{>0}_{\mathbf{v}_{+},\mathbf{w}}\times\overline{T}_{>0}\xrightarrow{\sim}\mathcal{S}_{\phi(v,w)}^{>0}. (3) For every (v,w)\in P , there is an isomorphism of semi-algebraic varieties H^{>0}_{\mathbf{v}_{+},\mathbf{w}}\times\overline{T}_{>0}\xrightarrow{\sim}\mathcal{S}_{\phi(v,w)^{-1}}^{>0}. (4) The space \mathcal{S}_{\geq 0} admits the cell decomposition \mathcal{S}_{\geq 0}=\bigsqcup_{u\in Q}\mathcal{S}_{u}^{>0}. (5) We have the closure relations \overline{\mathcal{S}_{u}^{>0}}=\bigsqcup_{u\leq u^{\prime}}\mathcal{S}_{u^{\prime}}^{>0}. (6) Let u\in Q . The cell \mathcal{S}_{u}^{>0} is a connected component of \mathring{\mathcal{S}}_{u}(\mathbb{R}) . (7) Let (v,w)\in P and write \sigma(v,w)=(v^{\prime},w^{\prime}) . After identifying the quotient torus \overline{T}_{>0} with positive real coordinates, the transition map \mathbb{R}_{>0}^{\dim\overline{T}+\ell(w)-\ell(v)}\rightarrow G^{>0}_{\mathbf{v}_{+},\mathbf{w}}\times\overline{T}_{>0}\rightarrow\mathcal{S}_{\phi(v,w)}^{>0}\rightarrow H^{>0}_{\mathbf{v}^{\prime}_{+},\mathbf{w}^{\prime}}\times\overline{T}_{>0}\rightarrow\mathbb{R}_{>0}^{\dim\overline{T}+\ell(w)-\ell(v)} is subtraction-free. (8) The total nonnegative symmetric space \mathcal{S}_{\geq 0} is contractible. 1.5. Lusztig [ LusztigTotalPositivitySymmetricSpaces ] has already introduced a notion of total positivity for symmetric spaces. Our approach is different. First, our pinning is based on a split pair (B^{+},T) , which is opposite to Lusztig’s choice of a fundamental pair. This leads to different definitions of totally nonnegative symmetric spaces. Second, we use an entirely different stratification: the double Bruhat cell stratification motivated by Poisson geometry. This leads to a particularly simple cell decomposition, closure order, and subtraction-free transition functions. In particular, after passing to the quotient torus, the transition functions are subtraction-free without taking square roots. 1.6. The paper is organized as follows. In §2 we recall the necessary background on symmetric spaces, the embedding G/K\hookrightarrow G , Lusztig total positivity, and Marsh–Rietsch parametrizations. In §3.1 we prove the poset isomorphism between the symmetric space indexing set and the projected Richardson indexing set. In §3.2 we define