Topological Phases of Matter in the Condensed-Mathematics Paradigm: A Modular Research Program

1 Introduction

1.1 The proposal

A quantum phase is usually presented as an entry in a classification table. This program takes a different starting point, due in outline to the seed prospectus of the series and realized in detail across Parts I–V: a phase is a component of a condensed moduli object of uniformly gapped local Hamiltonians. Fix a spatial dimension $d$, a lattice or coarse metric space $L$, on-site Hilbert spaces, a locality (interaction-decay) class, and an internal or spatial symmetry group $G$. Let $\mathcal{I}_{d,G}$ be the space of admissible $G$-symmetric local or quasi-local interactions. Rather than use $\mathcal{I}_{d,G}$ as an ordinary topological space, pass to its condensation $$X\longmapsto\underline{X},\qquad \underline{X}(S)=\operatorname{Cont}(S,X),\qquad S\ \text{profinite},$$ a sheaf on the site of profinite sets with finite jointly surjective covers. On compactly generated spaces the passage $X↦\underline{X}$ is fully faithful [1,2], so coupling tori, Banach interaction spaces, and compact disorder hulls are not discarded but embedded into a category with exact homological algebra. Because Hamiltonians carry gauge and quasi-local automorphisms, the right object is a condensed anima or stack, not a condensed set.

The whole program is the study of one geometric object and one substack. Over a profinite probe $S$ let $\mathfrak{Ham}_{d,G}(S)$ be the $S$-continuous families of $G$-symmetric quasi-local interactions, and let $$\mathfrak{Gap}_{d,G}(S)=\bigcup_{\Delta>0}\{\,H\in\mathfrak{Ham}_{d,G}(S):\mathop{\mathrm{inf}}_{s\in S}\operatorname{gap}(H_s)\ge\Delta\,\}$$ be the uniformly gapped substack (the word uniformly being load-bearing, since a family whose pointwise gaps degenerate to zero must be excluded). Inverting a class $\mathcal{W}$ of gapped adiabatic / quasi-local equivalences and stabilizing by trivial ancillas produces the phase $∞$-groupoid, and its shape records the phases: $$\mathfrak{Phase}_{d,G}:=\bigl(\mathfrak{Gap}_{d,G}[\mathcal{W}^{-1}]\bigr)^{\mathrm{st}}, \qquad \operatorname{Phases}_{d,G}=\pi_0\,\operatorname{Shape}(\mathfrak{Phase}_{d,G}).$$ Stacking systems, $\boxtimes$, makes $\mathfrak{Phase}_{d,G}$ symmetric monoidal; group-completing its invertible sector yields the connective condensed spectrum $\mathbf{IP}^{\mathrm{cond}}_{d,G}$. The two slogans of the program are the anchors of everything below: $$\boxed{\;\text{topological phase}=\text{a component of the stabilized condensed stack of gapped systems}\;}$$ and $$\boxed{\; \begin{array}{c} \text{topological transition}={}\\[2pt] \text{crossing the gapless discriminant in the Hamiltonian moduli stack}. \end{array} \;}$$

1.2 The program as a modular hierarchy

What Parts I–V supply is not a monolith but five interoperable modules, each of which takes the previous module’s output as its input and produces new structure. This is the sense in which the program is modular: the composition is a hierarchy, and at each level of the hierarchy a property emerges that was not visible one level down.

1. Locality (Part I [3]). Makes $\mathcal{I}_{d,G}$ a real Banach space $\mathcal{B}_{F}$ through a Nachtergaele–Sims–Young $F$-function locality norm, and promotes the Heisenberg dynamics to a morphism of condensed sets with a base-uniform Lieb–Robinson light cone. Emergent structure: the ground floor $\mathfrak{Ham}_{d,G}$ exists as a condensed object with a well-defined notion of quasi-local time evolution over any profinite base.

2. Positivity and $C^\ast$-norms (Part II [4]). Attaches the quasi-local $C^\ast$-algebra $\mathcal{A}$, its compact condensed state space, and, via Aoki’s solidification theorem, a solid $K$-theory invariant. Emergent structure: functorial topological invariants of disordered families over profinite hulls, produced by an intrinsically condensed operation.

3. The gap (Part III [5]). Cuts out the uniformly gapped substack $\mathfrak{Gap}_{d,G}$ and proves its stability where a stability theory holds with uniform constants. Emergent structure: a robust set of phases $\operatorname{Phases}_{d,G}$, once $\mathcal{W}$ is inverted and ancillas are added.

4. Stabilization and effective field theory (Part IV [6]). Group-completes the invertible sector into the connective condensed spectrum $\mathbf{IP}^{\mathrm{cond}}_{d,G}$ and compares it with the Freed–Hopkins bordism classification. Emergent structure: the higher homotopy of pumps and defects, and the relative charge $\partial\nu$ as a spectral boundary map.

5. Realizability (Part V [7]). Identifies which spectral classes are produced by an explicit gapped lattice Hamiltonian. Emergent structure: an obstruction subgroup separating abstract classes from lattice-realizable ones.

Two threads cut across all five modules and are what make the composition cohere rather than merely stack. The first is the transition calculus $(U_f,\Sigma_f,\nu_f,\partial\nu)$, which ties modules III–V together: the gapped locus of a family, its gapless discriminant, the locally constant phase label, and the relative charge of a crossing. The second is the profinite disorder thread $\Omega=F^{\mathbb{Z}^d}$, which runs through every module: descent of Lieb–Robinson estimates in Part I, crossed-product observable algebras in Part II, uniform-gap descent in Part III, parametrized invariants in Part IV, and functorial realizability in Part V.

1.3 What is proved and what is conjectured

The stance of this synthesis is the stance of the whole series: a rigorous research program, not a finished theory. We are strict about the boundary. The module-level results (Theorems I-A through V-C, and the citations of Aoki, Kubota, Freed–Hopkins, Ogata, and Kapustin–Fidkowski as rigorous anchors) are theorems, and we restate them (as Recollections, with citations to the Parts that prove them) rather than reprove them. The program-level claims are open. That the modules glue into a single condensed higher stack with $\mathbf{IP}^{\mathrm{cond}}_{d,G}$ as its invertible spectrum; that the microscopic and field-theoretic classifications agree; that the relative charge is the boundary map of that spectrum; that every program invariant descends along disorder hulls; each of these is stated below as a numbered Conjecture VI-$k$ and is never asserted as proven.

Two hard theorems bound the program from outside and are quoted wherever they bite. Cubitt–Pérez-García–Wolf [8] proved that the spectral gap is undecidable, so there is no algorithm and no uniform criterion deciding membership in $\mathfrak{Gap}_{d,G}$; existence of a thermodynamic gap is a hypothesis that defines the substack, not a computable property. Two commuting-projector no-gos bound realizability: Kapustin–Fidkowski [9] proved that a $U(1)$-symmetric local commuting-projector Hamiltonian has zero electric Hall conductance ($\sigma_H=0$), and Kapustin–Spodyneiko [10] proved that its chiral central charge vanishes ($c_-=0$) with no symmetry hypothesis; so every chiral invertible class (nonzero $\sigma_H$ under $U(1)$, or nonzero $c_-$) lies outside the commuting-projector image. Neither wall is ever crossed in what follows.

1.4 Relation to companion papers

This paper is the capstone of a six-part series and consumes the outputs of Parts I–V; it imports the canonical notation of the series verbatim and does not restate it. We summarize each companion in one paragraph, and Section 3 develops the composition in full.

Part I, Condensed Locality [3], is the analytic ground floor. Fixing an $F$-function it makes the $G$-symmetric interactions of finite $F$-norm a real Banach space $\mathcal{B}_{F}$, condenses it to $\underline{\mathcal{B}_{F}}(S)=\operatorname{Cont}(S,\mathcal{B}_{F})$, and proves that Heisenberg dynamics is a strongly continuous group of $\ast$-automorphisms obeying a Lieb–Robinson bound (Theorem I-A), that its assignment is a morphism of light condensed sets (Theorem I-B), and that a profinite family is exactly compatible finite data with a base-uniform light cone (Theorem I-C). It generates the equivalence class $\mathcal{W}$ from quasi-adiabatic continuation and states Conjectures I-1 (condensed higher stack), I-2 (condensed group of quasi-local automorphisms), and I-3 (descent of Lieb–Robinson estimates along a disorder hull).

Part II, Positivity, $C^\ast$-Norms, and Condensed State Spaces [4], is the observable side. It shows the state space $\mathcal{S}(\mathcal{A})$ condenses to a compact Hausdorff condensed set cut out by positivity and normalization (Theorem II-A), that Aoki’s solidification recovers operator $K$-theory after Bott inversion (Theorem II-B), and that Bellissard’s crossed product $C(\Omega)\rtimes\mathbb{Z}^d$ condenses functorially in the finite quotients of the hull (Theorem II-C). It records Conjectures II-1 through II-5 (closed condensed substack; naturality of the solid invariant; Real/$KKO$ refinement; solid classifying object; condensed superselection). Following Part II we write the disorder alphabet as $Q$, not $F$, to avoid collision with the $F$-function of Part I; see Section 8.

Part III, The Uniformly Gapped Substack [5], is the gap module. It proves that the finite-volume gap is Lipschitz and the gapped locus is open at finite resolution (Theorem III-A), that $\mathfrak{Gap}_{d,G}$ is condensed-open on the frustration-free / LTQO stratum (Theorem III-B), and that a uniform gap forces uniform clustering (Theorem III-C), all against the CPW undecidability wall; it states Conjectures III-1 (openness on the physical stratum), III-2 (uniform-gap descent), and III-3 (quasi-adiabatic continuation realizes $\mathcal{W}$).

Part IV, From Lattice Models to Effective Field Theories [6], is the stabilization module. It makes $\boxtimes$ a symmetric-monoidal structure, group-completes $\operatorname{Phases}_{d,G}$, assembles $\mathbf{IP}^{\mathrm{cond}}_{d,G}$ with Kubota’s $\Omega$-spectrum as its rigorous carrier, and sets the lattice classification against Freed–Hopkins. It states Conjectures IV-1 (condensed/solid refinement), IV-2 (lattice–EFT comparison equivalence), IV-3 (relative charge as spectral boundary map), IV-4 (solidification commutes with group completion), and IV-5 (renormalization functor).

Part V, Physical Realizability [7], is the realizability module. It proves that every group-cohomology class is realized by a commuting-projector model (Theorem V-A), completeness of the operator-algebraic index in $d=1$ for on-site finite symmetry, with the $d=2$ index surjective but its completeness open (Theorem V-B and Remark V-B$'$), and the two commuting-projector no-gos, electric (Kapustin–Fidkowski) and thermal/chiral (Kapustin–Spodyneiko) (Theorem V-C). It states Conjectures V-1 (realizability image is the SRE subspectrum), V-2 (chiral realizability by non-commuting models), and V-3 (profinite-family realizability).

1.5 Outline

Section 2 recalls the condensed-mathematics substrate and fixes conventions. Section 3 is the spine: the boxed sequence, arrow by arrow, with each module’s theorems restated and cited and its emergent structure named. Section 4 makes modular composition explicit: the composition diagram, the emergent properties, and the two walls. Section 5 states the homotopy dictionary. Section 6 develops the relative-charge formalism. Section 7 works the SSH chain through all five modules. Section 8 runs the profinite disorder thread. Section 9 states the program-level Conjectures VI-1 through VI-4 and consolidates all nineteen module conjectures with dependencies. Section 10 weighs what the paradigm contributes and what it does not, and Section 11 concludes.

2 The condensed-mathematics substrate

We recall only what the composition uses; Parts I and II give the careful development.

2.1 Condensation and profinite probes

A condensed set is a sheaf on the site of profinite sets $S$ with covers the finite jointly surjective families [1]; the pyknotic variant of Barwick–Haine [2] differs only in set-theoretic bookkeeping. The functor (1) embeds compactly generated topological spaces fully faithfully, so no information is lost in passing from $X$ to $\underline{X}$. The point of the passage is algebraic: condensed abelian groups form a Grothendieck abelian category with exact products and derived functors, where ordinary topological abelian groups do not, and this is precisely the environment in which derived invariants, descent, and completions behave. A condensed object is light when it is generated under the site by a countable family (the case for separable Banach spaces and separable $C^\ast$-algebras), which is why Parts I and II take care to work with separable data.

2.2 Stacks, anima, and shape

Because a Hamiltonian carries automorphisms (gauge transformations, on-site symmetry actions, quasi-local unitaries), the moduli problem $S↦\{S\text{-families of Hamiltonians}\}$ valued in groupoids (or $∞$-groupoids) is a candidate condensed stack, not merely a condensed set. Whether it satisfies descent along profinite covers is the content of Conjecture I-1 and is not proved. The shape $\operatorname{Shape}(-)$ of a condensed anima is its underlying homotopy type; $\pi_0\operatorname{Shape}$ extracts the ordinary set of components, hence $\operatorname{Phases}_{d,G}$ in (3). Throughout, fraktur denotes the moduli stacks ($\mathfrak{Ham},\mathfrak{Gap},\mathfrak{Phase}$), calligraphic the interaction space and the equivalence class ($\mathcal{I},\mathcal{W}$), boldface the spectra ($\mathbf{IP}^{\mathrm{cond}}$), and $\underline{(-)}$ condensation.

2.3 The boxed sequence

The program is summarized by one sequence, quoted from the prospectus and realized in Parts I–V: $$\boxed{ \begin{array}{c} \text{local quantum interactions}\\[2pt] \big\downarrow\ \text{ Part I}\\[2pt] \text{condensed moduli stack of Hamiltonians }\mathfrak{Ham}_{d,G}\\[2pt] \big\downarrow\ \text{ Part II}\\[2pt] \text{condensed observable algebras and solid }K\text{-theory}\\[2pt] \big\downarrow\ \text{ Part III}\\[2pt] \text{uniformly gapped substack }\mathfrak{Gap}_{d,G}\\[2pt] \big\downarrow\ \text{ Part IV}\\[2pt] \text{stabilized phase }∞\text{-groupoid }\mathfrak{Phase}_{d,G}\\[2pt] \big\downarrow\ \text{ Part IV}\\[2pt] \text{invertible condensed phase spectrum }\mathbf{IP}^{\mathrm{cond}}_{d,G}\\[2pt] \big\downarrow\ \text{ Part V}\\[2pt] \text{realized subspectrum inside }\mathbf{IP}^{\mathrm{cond}}_{d,G}. \end{array}}$$ The prospectus states the middle five lines; Parts II and V add the observable-algebra and realization refinements. Each arrow is a module. Section 3 treats them in turn.

3 The spine: the boxed sequence and its five modules

We walk down (6). For each module we recall its principal theorems (as Recollections, attributed to the Part that proves them) and name the structure that emerges. Nothing in this section is claimed as new; the novelty of the synthesis is the composition, treated in Section 4.

3.1 Arrow I: local interactions to the condensed stack

The first arrow turns a physicist’s list of local couplings into a condensed object that can be probed by profinite sets. Part I fixes an $F$-function $F$ encoding interaction decay and defines the interaction Banach space $\mathcal{B}_{F}$.

Emergent structure. The ground floor $\mathfrak{Ham}_{d,G}$ is a condensed object over which time evolution is quasi-local uniformly in the probe. Uniformity is what makes every later module’s “family over a base” well-posed: it is the reason a disordered family has a single light cone, a single clustering length, a single set of constants. Part I’s Conjectures I-1, I-2, I-3 mark the edge: stack descent, a condensed automorphism group, and Lieb–Robinson descent along a hull; none proved.

3.2 Arrow II: observable algebras and solid $K$-theory

The second arrow attaches to the stack the algebra of observables and the $K$-theory from which topological invariants are built. For a spin system with finite on-site dimension $\mathcal{A}$ is a separable unital AF algebra (UHF in the homogeneous case), and its condensation is a sheaf of $C^\ast$-algebras, light because $\mathcal{A}$ is separable.

The load-bearing bridge is Aoki’s theorem, and the honesty of the whole program depends on stating it with its Bott-inversion caveat, exactly as Part II now does.

We stress the bookkeeping because it is easy to get wrong and consequential when composed. Writing $K_{\mathrm{op}}(A)\simeq\operatorname{Solid}(K_{\mathrm{alg}}(\underline{A}))$ without the inversion $[\beta^{-1}]$ is false in negative degrees: solidification of connective algebraic $K$-theory lands in a connective spectrum with no negative homotopy, whereas operator $K$-theory is periodic. The two agree in nonnegative degrees, which is all the SSH winding (a degree-$\ge 0$ class) needs, but the periodic identification requires (7). Every later use of “the solid invariant” in this synthesis carries this qualification.

Emergent structure. The composition of Arrows I and II produces functorial topological invariants of disordered families over profinite hulls, computed by an intrinsically condensed operation (solidification) rather than imported by hand. This is the first place the condensed language does work an ordinary smooth-parameter treatment does not: the profinite probes of (1) match the physical configuration space $\Omega$ of Recollection 3.5, they do not approximate it. Part II’s Conjectures II-1 through II-5 record what is open here.

3.3 Arrow III: the uniformly gapped substack

The third arrow selects the systems on which a phase invariant is even defined. The selection is governed by a wall.

Consequently membership in $\mathfrak{Gap}_{d,G}$ is not, in general, decidable, and Part III does not attempt to decide it. Existence of a thermodynamic gap defines the substack; the content is stability.

Emergent structure. Inverting $\mathcal{W}$ and stabilizing on $\mathfrak{Gap}_{d,G}$ yields a robust $\pi_0$: a set of phases stable under quasi-local deformation, disorder averaging, and finite-depth circuits. The robustness is exactly what Recollections 3.7 and 3.8 buy: openness so a phase label does not change under small perturbations, clustering so it is detected locally. But the openness holds only on the stratum where the stability hypotheses hold; Conjecture III-1 asks for it on the full physical stratum, III-2 for descent along hulls, III-3 for the identification of $\mathcal{W}$-components with Ogata’s operator-algebraic phases [19].

3.4 Arrow IV: stabilization and the invertible condensed phase spectrum

The fourth arrow group-completes the invertible sector into a spectrum. On components $\boxtimes$ makes $\operatorname{Phases}_{d,G}$ a commutative monoid with unit the trivial product state; the short-range-entangled (SRE) phases are exactly its invertible elements.

Applying the recognition principle for grouplike $E_∞$-spaces to the invertible sector yields a connective spectrum, the invertible condensed phase spectrum $\mathbf{IP}^{\mathrm{cond}}_{d,G}$ [22]. Against it Part IV sets the effective-field-theory target: the Freed–Hopkins Anderson-dual bordism spectrum $I_{\mathrm{FH}}$ classifying invertible topological field theories [23,24].

Emergent structure. Two things appear only at the spectrum level. First, the higher homotopy: $\pi_1$ of the phase object is loops of gapped Hamiltonians, which implement adiabatic pumps and automorphisms of topological order [28]; $\pi_n$ is $n$-parameter families and higher defects. Second, the transition calculus becomes a long-exact-sequence computation, the relative charge $\partial\nu$ appearing as a boundary map (Section 6). Part IV’s Conjectures IV-1 through IV-5 govern the condensed refinement, the lattice–EFT comparison, the boundary-map identity, the compatibility of solidification with group completion, and the renormalization functor.

3.5 Arrow V: realizability

The fifth arrow asks the converse question: which classes in $\mathbf{IP}^{\mathrm{cond}}_{d,G}$ are produced by an explicit uniformly gapped lattice Hamiltonian? Part V formalizes realizability as essential surjectivity on $\pi_0$ of the comparison map $\operatorname{real}$ from the stack of lattice models to the abstract spectrum, and separates realizability by some gapped local model from realizability by a local commuting-projector model.

Emergent structure. The composition of all five arrows produces an obstruction subgroup $\operatorname{Obs}_{d,G}=\pi_0\mathbf{IP}^{\mathrm{cond}}_{d,G}/\operatorname{im}\operatorname{real}$ separating abstract classes from lattice-realizable ones, with the chiral classes of Recollection 3.14 sitting inside the commuting-projector obstruction. Part V’s Conjectures V-1, V-2, V-3 describe the image, the chiral classes, and profinite-family realizability.

4 Modular composition

The previous section walked the arrows in isolation. Here we make the composition explicit: how the modules compose, what emerges from composition that is invisible in any single module, and what bounds the composition from outside.

4.1 The composition diagram

Read as modules, the five Parts form a diagram in which each object is the input to the next and each module contributes a functor. Write $\mathsf{Loc}$, $\mathsf{Obs}$, $\mathsf{Gap}$, $\mathsf{Stab}$, $\mathsf{Real}$ for the five modules.

\begin{tikzcd}[column sep=1.5em, row sep=2.6em]
\{\text{loc.\ int.}\}
  \arrow[r, "\mathsf{Loc}"]
& \mathfrak{Ham}_{d,G}
  \arrow[r, "\mathsf{Gap}"]
  \arrow[d, "\mathsf{Obs}"']
& \mathfrak{Gap}_{d,G}
  \arrow[r, "\mathsf{Stab}"]
& \mathfrak{Phase}_{d,G}
  \arrow[r, "\boxtimes\text{, }\operatorname{Groth}"]
& \mathbf{IP}^{\mathrm{cond}}_{d,G}
  \arrow[d, "c_{d,G}"]
  \arrow[dl, dashed, "\operatorname{real}"'] \\
& (\mathcal{A},\ \underline{\mathcal{S}(\mathcal{A})},\ \operatorname{Solid}K_{\mathrm{alg}})
&& \{\text{realized}\}
& I_{\mathrm{FH}}
\end{tikzcd}

Here $\mathsf{Loc}$, $\mathsf{Obs}$, $\mathsf{Gap}$, $\mathsf{Stab}$, and the group-completion arrow are the functors of Parts I, II, III, III–IV, and IV; the solid arrows are functors the modules construct, with the caveat that $\mathsf{Stab}$ inverts $\mathcal{W}$, whose generation from quasi-adiabatic continuation is Conjecture III-3. The dashed arrow $\operatorname{real}$ is the realization comparison of Section 3.5 (Part V), whose essential image is Conjecture V-1; the vertical comparison $c_{d,G}:\mathbf{IP}^{\mathrm{cond}}_{d,G}\to I_{\mathrm{FH}}$ to the Freed–Hopkins target is Conjecture IV-2. The observable data $(\mathcal{A},\underline{\mathcal{S}(\mathcal{A})},\operatorname{Solid}K_{\mathrm{alg}})$ produced by $\mathsf{Obs}$ feeds the homotopy of the spectrum through the solid refinement of Conjecture IV-1. The diagram commutes at the level of $\pi_0$ exactly where those conjectures hold—in $d=1$ for on-site finite symmetry (Recollection 3.13), and for the free-fermion tenfold way [21,33]—and its global commutativity is the Master Conjecture of Section 9.

4.2 Emergent properties at each level

The characteristic feature of a modular composition is that structure emerges at each level which is not a property of the parts. We collect the four emergences named in Section 3.

1. Uniformity (I$\circ$nothing). A single light cone, clustering length, and constant set over a whole profinite base: the precondition for every “family” below.

2. Functorial disordered invariants (II$\circ$I). Topological invariants of disordered families computed by solidification, matched to the physical configuration space rather than a smooth approximation of it.

3. A robust phase set (III$\circ$II$\circ$I). $\operatorname{Phases}_{d,G}=\pi_0\operatorname{Shape}\mathfrak{Phase}_{d,G}$, stable under quasi-local deformation once $\mathcal{W}$ is inverted and ancillas added.

4. Higher homotopy and an obstruction subgroup (V$\circ$IV$\circ$III$\circ$II$\circ$I). Pumps and defects as $\pi_{\ge1}$, and $\operatorname{Obs}_{d,G}$ separating abstract from realizable.

None of these emergences is forced by the module below it alone; each requires the composite. That is the precise content of calling the program modular rather than layered: the modules interoperate, and the interoperation is where the physics of families, disorder, and defects lives.

4.3 The two walls

Two theorems bound the composition and must be respected at every level.

The undecidability wall (Recollection 3.6) forbids any claim that $\mathfrak{Gap}_{d,G}$ is decidable or algorithmically presentable. It does not forbid stability results: openness on a stratum, uniform clustering, and descent along a hull are all statements about the neighborhood of a gapped system, not tests for gappedness, and Part III is careful to scope them to strata with uniform constants. The wall is why the program is organized around stability, not existence.

The no-go wall (Recollection 3.14), combining the electric Kapustin–Fidkowski and the thermal Kapustin–Spodyneiko no-gos, forbids any claim that a chiral invertible class is realized by commuting projectors. It bounds $\operatorname{im}\operatorname{real}^{\mathrm{cp}}$ to the non-chiral sector and makes the separation between “exactly solvable” and “physically realizable” structural, not incidental. Every realizability statement in the program excludes the chiral commuting-projector case by hypothesis.

A third boundary is not a theorem but a scope limit: a spectrum classifies only invertible order. General anyon theories and non-invertible phases need higher categories of excitations (braided or modular tensor categories in $2{+}1$ dimensions and their higher analogues), so a condensed treatment of them would require a condensed higher stack of phases and defects, not a condensed $K$-theory spectrum. That extension is outside the present program.

5 The homotopy dictionary

Classification by $\pi_0$ records only which phases exist. The higher homotopy of the phase object carries more, and the program fixes a dictionary between homotopy-theoretic and physical data. We quote it verbatim from the prospectus, now with each line anchored to the module that supplies it.

$$\boxed{ \begin{aligned} \pi_0 &= \text{phases},\\ \pi_1 &= \text{adiabatic pumps and phase automorphisms},\\ \pi_n &= \text{higher families and higher defects},\\ \Sigma&= \text{gapless transition locus},\\ E^{q+1}(B,B\setminus\Sigma) &= \text{relative charge of a transition}. \end{aligned}}$$

The first line is Arrow III’s emergent structure, made precise by Conjecture III-3: the components of $\mathfrak{Gap}_{d,G}[\mathcal{W}^{-1}]$ are the operator-algebraic gapped phases of Ogata [19]. The second and third are Arrow IV’s: loops and higher families of gapped Hamiltonians realize pumps and defects, carried by the homotopy of Kubota’s spectrum [20] and, conjecturally, by the higher homotopy solid modules of $\mathbf{IP}^{\mathrm{cond}}_{d,G}$ (Conjecture IV-1); the automorphism content of $\pi_1$ is the subject of [28], whose $\pi_1,\pi_2,\pi_3$ computations are themselves conjectural. The last two lines are the transition calculus, developed next.

6 Phase transitions and the relative-charge formalism

This section develops the fourth and fifth lines of (9): what a transition is, and what its charge is. The development is honest obstruction theory (the long exact sequence of a pair and excision, valid for any generalized cohomology), and the only conjectural step is the identification of the abstract boundary map with the spectral boundary map of $\mathbf{IP}^{\mathrm{cond}}_{d,G}$, which is Conjecture IV-3 lifted to the program.

6.1 The gapped locus, the discriminant, and the phase label

Let $B$ be a parameter space (couplings, fields, pressures, disorder configurations, boundary conditions), regarded as a condensed object $\underline{B}$, and let $f:\underline{B}\to\mathfrak{Ham}_{d,G}$ be a family of systems.

That $\nu_f$ is locally constant on $U_f$ is exactly Recollection 3.7: the gapped locus is condensed-open on the physical stratum, and openness of each phase’s preimage is what “locally constant” means. A topological phase transition is a path in $B$ whose endpoints carry different values of $\nu_f$; such a path cannot stay in $U_f$ and must meet $\Sigma_f$. This is the boxed slogan (5).

6.2 The relative charge as an obstruction

Suppose the invariant is valued in a (condensed) generalized cohomology theory $E$; in the invertible case $E$ is represented by $\mathbf{IP}^{\mathrm{cond}}_{d,G}$. On the gapped locus the invariant is a class $\nu\in E^q(U_f)$. The question of the transition is whether $\nu$ extends over the critical set to all of $B$. Obstruction theory answers it through the long exact sequence of the pair $(B,U_f)$: $$·s\to E^q(B)\xrightarrow{\ j^\ast\ }E^q(U_f)\xrightarrow{\ \partial\ } E^{q+1}(B,U_f)\xrightarrow{\ \ }E^{q+1}(B)\to·s,$$ with $j:U_f\hookrightarrow B$ the inclusion.

6.3 Locality: linking spheres

The relative group $E^{q+1}(B,B\setminus\Sigma_f)$ is local along $\Sigma_f$. When $\Sigma_f$ is a reasonable (say, closed, locally-flat, finite-codimension) condensed subobject with components $\{\Sigma_f^{(\alpha)}\}$, excision identifies the relative cohomology with a sum of contributions supported near the components: $$E^{q+1}(B,B\setminus\Sigma_f)\ \cong\ \bigoplus_\alpha E^{q+1}\bigl(N_\alpha,N_\alpha\setminus\Sigma_f^{(\alpha)}\bigr),$$ $N_\alpha$ a tubular neighborhood of $\Sigma_f^{(\alpha)}$. If $\Sigma_f^{(\alpha)}$ has codimension $c$ with an $E$-oriented normal bundle, the Thom isomorphism identifies each summand with a degree-shifted absolute group of the component, $$E^{q+1}\bigl(N_\alpha,N_\alpha\setminus\Sigma_f^{(\alpha)}\bigr)\ \cong\ E^{q+1-c}\bigl(\Sigma_f^{(\alpha)}\bigr),$$ the Thom class shifting degree by the codimension $c$. The local charge is read off on a linking sphere: over a point $x\in\Sigma_f^{(\alpha)}$ the unit sphere $S^{c-1}_{\mathrm{link}}$ of the normal fibre lies in $B\setminus\Sigma_f$ and links the component, and the restriction $\nu|_{S^{c-1}_{\mathrm{link}}}\in E^{q}\bigl(S^{c-1}_{\mathrm{link}}\bigr)$ has reduced part $\widetilde E^{\,q}(S^{c-1})\cong E^{q-c+1}(\mathrm{pt})$ equal to that local charge: the same degree shift as (12). In words: the relative charge decomposes as a sum of local charges, one per component of the critical set, each detected by evaluating $E$ on a small sphere linking the component. This is the precise form of the physical picture that a band degeneracy (a Dirac node, a Weyl point) acts as a source or sink of topological charge, now valid for a generalized cohomology theory and an interacting phase spectrum rather than only for Chern numbers of free bands. This relative, defect-localized reading has a free-fermion precedent: Teo and Kane classify topological defects by the $K$-theory of a sphere linking the defect [34], within the $K$-theoretic classification of topological phases [35]; the condensed formulation of Definition 6.2 is the extension of that relative-charge picture to a generalized theory represented by $\mathbf{IP}^{\mathrm{cond}}_{d,G}$. The low-codimension case $c=1$ (a wall $\Sigma_f$ separating two gapped regions) reduces the linking sphere to $S^0$: two points, one in each phase, and the local charge is the difference of the two phase labels: the jump of $\nu_f$ across the wall.

6.4 The spectral boundary map

Proposition 6.3 and (12) are unconditional facts about any $E$. The program’s substantive claim is that, for $E=\mathbf{IP}^{\mathrm{cond}}_{d,G}$-cohomology, the connecting map $\partial$ of (10) is the boundary map of the cofiber sequence of the pair for the spectrum $\mathbf{IP}^{\mathrm{cond}}_{d,G}$, so the transition charge is computed by a single long exact sequence in every dimension and every generalized theory simultaneously. This is Conjecture IV-3, restated at the program level as Conjecture VI-3 in Section 9. Its low-dimensional shadow (the SSH winding jump, the Dirac node as a unit source of charge) is classical; the conjecture is that this is systematically the spectral boundary map.

7 The SSH transition end to end

We now run one example through all five modules, to show that the composition is not merely formal. The Su–Schrieffer–Heeger chain [36] is the smallest system in which every module has something to say, and its winding number is a degree-$\ge0$ class, so it sits safely on the nonnegative side of the Bott-inversion caveat of Recollection 3.4.

Now the five modules, in order.

Part I (dynamics). The SSH couplings define a finite-range interaction of finite $F$-norm, so the family lives in $\underline{\mathcal{B}_{F}}$ and, by Recollection 3.2, over any compact region of the $(t_1,t_2)$-plane away from $\Sigma$ carries one Lieb–Robinson velocity. The parameter torus and Brillouin circle are replaced by their condensations $\underline{B},\underline{S^1}$; the light cone is uniform over the probe.

Part II (algebra and $K$-theory). SSH is a class $\mathit{AIII}$ system in $d=1$, whose tenfold-way entry is $\mathbb{Z}$ [21,33], the winding number valued in $K^1$ of the observable algebra. Because the winding is a nonnegative-degree class, Recollection 3.4 applies without invoking the Bott inversion: the solid invariant $\operatorname{Solid}(K_{\mathrm{alg}}(\underline{A}))$ computes it directly, and its provenance is condensed even though the number is the classical winding.

Part III (gap). Proposition 7.2 is exactly the finite-volume/thermodynamic gap statement of Recollection 3.7 for this family: the gapped locus is $U=\{\,\lvert t_1 \rvert\ne\lvert t_2 \rvert\,\}$, open, and the phase label $\nu$ is locally constant on it, with the discriminant $\Sigma=\{\,\lvert t_1 \rvert=\lvert t_2 \rvert\,\}$.

Part IV (stabilization and EFT). The low-energy theory at the gap-closing $\lvert t_1 \rvert=\lvert t_2 \rvert$, $k=\pi$ is a massive $1{+}1$-dimensional Dirac fermion, and the winding number is the sign of the Dirac mass—the deformation class of the Dirac effective field theory. So the comparison map $c_{1,\mathit{AIII}}$ sends the SSH phase to its EFT class and the two agree: Conjecture IV-2 holds at the level of $\pi_0$ in this box.

Part V (realizability). The nontrivial SSH phase is short-range-entangled and realized by the explicit dimerized chain; being non-chiral ($\sigma_H=0$ and $c_-=0$), it is excluded by neither no-go of Recollection 3.14 and is in fact commuting-projector realizable in the flat-band (fully dimerized) limit. It sits in $\operatorname{im}\operatorname{real}$.

The transition charge. In the parameter plane $B=(t_1,t_2)$ the invariant relevant to the transition calculus is not the momentum-space $K^1$ class of the tenfold way but the winding number it produces: a locally constant $\mathbb{Z}$-valued label on the gapped locus, an element $\nu\in K^0(U)$ of degree $q=0$. The discriminant $\Sigma=\{\,\lvert t_1 \rvert=\lvert t_2 \rvert\,\}$ is the pair of lines $t_1=\pm t_2$; away from their intersection it has codimension $c=1$, so the linking sphere of (12) is $S^0$ and the local charge is read in the reduced group $\widetilde K^0(S^0)\cong\mathbb{Z}$—the winding difference $\pm1$ across the wall, in agreement with the degree shift $E^{q-c+1}(\mathrm{pt})=K^0(\mathrm{pt})=\mathbb{Z}$ of (12). (At the origin $(0,0)$ the two lines cross, the gap closes at $k=0$ and $k=\pi$ simultaneously, and the codimension jumps to $2$; the generic crossing is the codimension-one one just described.) By Proposition 6.3 the relative charge $\partial\nu\in K^{1}(B,U)$ is nonzero precisely because the winding label does not extend across $\Sigma$: the transition is real, and its charge is the unit jump. Conjecture VI-3 is the assertion that this $\partial$ is the spectral boundary map of $\mathbf{IP}^{\mathrm{cond}}_{1,\mathit{AIII}}$; here it is the elementary winding jump, and the two agree.

8 Disorder and the profinite thread

The profinite disorder thread is where condensed mathematics does work that ordinary smooth-parameter topology cannot, and it runs through all five modules. We collect it.

8.1 The hull as a profinite probe

If $Q$ is a finite set of local configurations, the configuration space $$\Omega=Q^{\mathbb{Z}^d}$$ is compact, totally disconnected, hence profinite; this is the disorder (or tiling) hull of Bellissard’s noncommutative geometry of aperiodic media [13,37]. (We follow Part II in writing the alphabet $Q$ rather than the prospectus’s $F$, to avoid collision with the Nachtergaele–Sims–Young $F$-function of Part I; the two never denote the same object.) A disordered Hamiltonian family $\omega↦ H_\omega$ is literally an $\Omega$-point of the moduli stack, $H\in\mathfrak{Ham}_{d,G}(\Omega)$, and finite quotients of $\Omega$ are finite-resolution disorder data. The sheaf condition of (1) is the compatibility of families and invariants across finite approximations, so the profinite probes match the physical configuration space rather than repackage a smooth one.

8.2 The thread through the modules

• Part I (Conjecture I-3): the Lieb–Robinson estimates over $\Omega$ should be the right Kan extension of their restrictions to finite quotients: quasi-local dynamics determined by finite-resolution data. Its unconditional shadow is the uniform light cone of Recollection 3.2.

• Part II (Recollection 3.5, Conjecture II-2): the observable algebra is the crossed product $C(\Omega)\rtimes_T\mathbb{Z}^d$, its condensation functorial in finite quotients, and the solid invariant should agree with operator $K$-theory naturally in $\Omega$.

• Part III (Conjecture III-2): uniform gappedness over $\Omega$ should be a closed condition on the inverse system: the uniformly gapped families the inverse limit of the finite-resolution ones at a fixed $\Delta$.

• Part IV: parametrized invariants over $\Omega$, the base level (L2) of the lattice–EFT comparison Conjecture IV-2.

• Part V (Conjecture V-3): realizability over $\Omega$ is disorder-robust iff a condensed-cohomological descent obstruction $o(c)\in H^1_{\mathrm{cond}}(\Omega,\underline{\operatorname{Aut}}(c))$ vanishes, detected by finite quotients.

The common shape of all five is descent along the finite quotients of $\Omega$. That the five descent statements are facets of one principle is Conjecture VI-4.

9 The program-level conjectures

We now state the open claims of the program. Each is built as the coherent join of module-level conjectures from Parts I–V, and none is proved. We write them as named Conjectures VI-1 through VI-4; the first is the Master Conjecture.

9.1 Consolidated open problems

Table 1 lists the nineteen module-level conjectures of Parts I–V, with the module each depends on and the program-level conjecture it feeds. The dependency column records which earlier module’s output a conjecture presupposes; the “feeds” column records which Conjecture VI-$k$ consolidates it. The table is the program’s to-do list.

Two features of Table 1 are worth naming. First, the dependency graph is a tree rooted at I-1: every conjecture presupposes the stack structure of the ground floor, which is why Conjecture I-1 is the true foundation and the Master Conjecture VI-1 leads with it. Second, the “feeds” column shows the program-level conjectures are not independent wishes but bundles: VI-1 gathers the ten structural conjectures, VI-2 the four comparison/refinement conjectures, VI-3 the single boundary-map conjecture, and VI-4 the four descent conjectures ($10+4+1+4=19$). Proving any program-level conjecture means proving its whole bundle.

10 Contributions and limitations

10.1 Contributions

The framework contributes three structural advantages, none of which is a new numerical invariant. First, one category holds ordinary continuous families, profinite disorder families, compact inverse limits, and topological symmetry groups at once; the passage (1) is fully faithful, so nothing classical is lost. Second, condensed abelian groups and solid modules supply exact derived operations and descent where ordinary topological groups and modules are poorly behaved: the environment Recollection 3.4 needs to make operator $K$-theory a condensed invariant. Third, the sheaf and stack viewpoint organizes local families, defects, interfaces, boundary conditions, symmetry actions, and gluing uniformly, which is what the transition calculus of Section 6 and the disorder thread of Section 8 exploit.

On novelty we are precise, because the reviewer should be. Condensed and pyknotic mathematics [1,2] and the solidification bridge [12] are established in pure mathematics; the moduli/space-of-states viewpoint on gapped systems is established in physics [20,22,38–40]. What is new is the synthesis: no prior work applies condensed or pyknotic machinery to topological phases, and the individual analytic ingredients (Lieb–Robinson bounds, $C^\ast$-positivity, gap stability, stabilization, realizability) are not new in themselves. The value is unificatory and organizational, and the honest content is the Master Conjecture and its bundle.

10.2 What it does not do

By itself the paradigm does not solve the hard analysis of many-body theory: one still needs locality estimates (Part I), positivity and $C^\ast$-norm conditions (Part II), the existence and stability of a thermodynamic gap (Part III), the lattice–EFT correspondence (Part IV), and physical realizability (Part V), and each is a genuine theorem or a genuine conjecture, not a corollary of the formalism. It does not decide the gap: the undecidability wall (Recollection 3.6) is permanent. It does not realize chiral phases by commuting projectors: the no-go wall (Recollection 3.14) is permanent. It does not, by construction, classify non-invertible topological order, which needs a condensed higher stack of phases and defects rather than a spectrum. And it does not produce a different Chern number: the winding of Proposition 7.2 is the classical winding; only its provenance is condensed. The genuinely new thesis is the single environment, and the genuinely open mathematics is Conjectures VI-1 through VI-4.

11 Conclusion

The six papers of this series are one program with five modules and a spectrum-level payoff. Part I makes the interaction space condensed and the dynamics a base-uniform morphism; Part II attaches observable algebras, a compact condensed state space, and the solid $K$-invariant with its Bott-inversion bookkeeping; Part III cuts out the uniformly gapped substack and stabilizes it against the undecidability wall; Part IV group-completes the invertible sector into $\mathbf{IP}^{\mathrm{cond}}_{d,G}$ and confronts the field-theoretic classification; Part V asks which classes a lattice realizes, bounded by the commuting-projector no-gos of Kapustin–Fidkowski and Kapustin–Spodyneiko. Composed, the modules give the boxed sequence (6), the homotopy dictionary (9), and the transition calculus in which a transition is the crossing of the discriminant and its charge is the relative class $\partial\nu$ localized by linking spheres.

We have been careful to keep the ledger honest. The module results are theorems, cited to Parts I–V. The program results (that the modules glue into one condensed higher stack with $\mathbf{IP}^{\mathrm{cond}}_{d,G}$ its invertible spectrum, that microscopic and field-theoretic classifications agree, that the relative charge is a spectral boundary map, that all invariants descend along disorder hulls) are Conjectures VI-1 through VI-4, and Table 1 records exactly what each would need. The program is not a completed classification and does not claim to be. It is a single categorical and homological environment in which continuous families, profinite disorder, analytic completions, operator $K$-theory, symmetry, stacking, defects, and phase-transition loci can be treated at once, and a precise list of what remains to prove.

Code availability

The formal-verification suites accompanying the six-paper series are at github.com/YonedaAI/topological-phases-of-matter; each module’s code lives in a per-topic subdirectory (for example src/bordism-realizability/ and src/lattice-eft-equivalence/). This synthesis introduces no new code of its own; it composes the results verified in Parts I–V.

Acknowledgements

This is Part VI of a six-part series by the author and The YonedaAI Collaboration; it depends entirely on the analytic and topological content of Parts I–V and on the foundational work cited throughout.

References