Uniqueness of the hyperspaces $C(p,X)$ in the class of trees

Florencio Corona-V\'azquez; Javier S\'anchez-Mart\'inez; Rosemberg Toal\'a-Enr\'iquez; Russell Aar\'on Qui\~nones-Estrella

REVIEW 3 major objections 5 minor 16 references

Uniqueness of the hyperspaces $C(p,X)$ in the class of trees

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read For every tree X and point p, the hyperspace C(p,X) of subcontinua containing p determines the pointed tree (X,p) among all trees.

desk verdict New and likely correct uniqueness theorem for pointed trees in C(p,X), but Section 4 silently switches from open cells to closures, leaving a repairable gap in the induction. read the letter →

arxiv 1908.06202 v1 pith:IYDDMRG6 submitted 2019-08-16 math.GN

Florencio Corona-Vázquez , Russell Aarón Quiñones-Estrella , Javier Sánchez-Martínez , Rosemberg Toalá-Enríquez This is my paper · ORCID

classification math.GN MSC 54B0554B2054F65

keywords hyperspacesC(pX)treescontinuauniquehyperspacecelldecompositionVietoristopologyhomogeneitydegree

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proves that pointed hyperspaces completely determine pointed trees. For any tree $X$, any point $p\in X$, and any other tree $Y$ with $q\in Y$, if the hyperspace $C(p,X)$ of all subcontinua of $X$ containing $p$ is homeomorphic to $C(q,Y)$, then there is a homeomorphism $h:X\to Y$ with $h(p)=q$. Earlier examples show the same statement fails when $Y$ is allowed to be a non-tree continuum, and fails for arcs versus simple closed curves inside the class of finite graphs, so the tree class is exactly the setting where this recovery works. The proof reads the graph structure of $X$ off the faces of the cells forming $C(p,X)$: components of its locally Euclidean part are indexed by subtrees of a trimmed core $T(X)$, and adjacent subtrees correspond to codimension-one faces. A direct corollary is that the number of homeomorphism classes among the hyperspaces $C(x,X)$ of a tree equals the homogeneity degree of $X$.

What carries the argument

The load-bearing object is the collection of open cells $U_G$ inside $C(p,X)$, indexed by subtrees $G$ of the trimmed core $T(X)$ (the tree obtained by deleting the end-point-incident edges of $X$). For a subtree $G$ containing $p$, $U_G$ consists of all continua in $C(p,X)$ that contain $G$ and extend a little way into each edge of $X\setminus G$ incident with $G$; each $U_G$ is homeomorphic to a product of open intervals, hence is a cell of dimension equal to the number of such incident edges. The key identity is Proposition 4.8: if $G'=G\cup e$ for an edge $e$, then the closures of $U_G$ and $U_{G'}$ meet in a nonempty face of codimension one, and conversely any nonempty codimension-one intersection of two such cells forces $G'=G\cup e$. Since homeomorphisms preserve intersections and dimensions, this face-incidence combinatorics is invariant, and the induction in Proposition 4.10 reconstructs every path from $p$ to a vertex in $T(X)$.

What would settle it

Compute, for a concrete tree $X$ with a ramification point $p$, the incidence graph whose vertices are the closures of the components $U_G$ of $U(X)$ and whose edges mark nonempty intersections of codimension one, and check whether a hypothetical homeomorphism $C(p,X)\cong C(q,Y)$ must preserve this graph. A pair of subtrees $G,G'\subset T(X)$ with $U_G$ meeting $U_{G'}$ in a set of dimension $\dim(U_G)-1$ while $G'\neq G\cup e$ would directly refute Proposition 4.8, and with it the reconstruction of paths in Proposition 4.10.

Watch

Extended reading notes

Core claim

The central discovery, Theorem 4.14, is that the map $(X,p)\mapsto C(p,X)$ is injective on pointed trees up to homeomorphism: if $C(p,X)\cong C(q,Y)$ with $X,Y$ trees, then $(X,p)\cong(Y,q)$. The argument first settles the easy cases (arcs and simple $n$-ods) using known classifications of their hyperspaces. For a tree with a ramification point $p$, Proposition 4.10 shows that a homeomorphism $h:C(p,X)\to C(q,Y)$ sends each cell component $U_{pp'}$ corresponding to the path from $p$ to a vertex $p'$ of the trimmed core $T(X)$ to the analogous component $U_{qq'}$ in $Y$, preserving the length of the path. The proof then assembles these paths into a bijection between vertex sets that preserves edges, and counts incident end-edges to extend the isomorphism to all vertices of $X$ and $Y$. Thus a homeomorphism of hyperspaces is promoted to an isomorphism of trees.

Load-bearing premise

The proof depends on the premise that the open components $U_G$ can be treated as closed cells whose face intersections have stable dimension and are preserved by any homeomorphism of $C(p,X)$; if a homeomorphism could distort that cell geometry, the reconstruction of the tree's edges from codimension-one face intersections would fail.

Editorial extensions

If this is right

Two pointed trees with homeomorphic pointed hyperspaces cannot differ: the homeomorphism lifts to a tree isomorphism sending the marked point to the marked point.
The order of the point $p$ in $X$ and the number of endpoints $|E(X)|$ are encoded in $C(p,X)$ as the minimal and maximal dimensions of its cells.
For a tree $X$, the size of $K(X)$, the number of homeomorphism classes among the hyperspaces $C(x,X)$ as $x$ ranges over $X$, equals the homogeneity degree of $X$.
The uniqueness statement is sharp within trees: the paper's examples show that allowing $Y$ to be a non-tree continuum, or allowing both trees to range over all finite graphs, destroys uniqueness.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The proof's mechanism suggests that the face-incidence graph of the cell decomposition of $C(p,X)$ is a homeomorphism invariant, so the hyperspace may determine not only the tree but also its combinatorial cell structure.
The argument uses finiteness of the edge set at every step; the authors' open question on dendrites would therefore need a different invariant, since infinite branching can produce infinite-dimensional cells.
Corollary 4.15 concerns the quotient $K(X)/\sim$; a natural strengthening to test is whether the equivalence classes of points under homeomorphism of $C(p,X)$ coincide with the orbits of the automorphism group of the tree.
One could test the sharpness of the theorem by asking whether uniqueness survives when $C(p,X)$ is replaced by the unpointed hyperspace $C(X)$ or by a quotient of $C(p,X)$; the paper does not address these variants.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Main take: the result—every pointed tree (X,p) is determined by C(p,X) inside the class of trees—is new and probably true, but the proof as written has a gap exactly where the induction needs it. I'd still send this to a referee; the repair is visible.

What is genuinely new: the uniqueness theorem is not in the cited literature, and the examples in Section 1 are useful because they show the class of all continua and even finite graphs fail, so the tree result is a real boundary. The cell decomposition idea is right, and the counting lemmas in Section 3, especially Corollaries 3.2–3.4, are clean. Corollary 4.15, which partially answers Problem 5.4 from the authors' earlier paper, is a legitimate application. No code or data are involved, but that is normal for this kind of proof.

The soft spot is real and load-bearing. Section 3 defines U_Y as an open subset of C(p,X) and proves the U_Y are pairwise disjoint components. Section 4 then writes U_G ∩ U_{G'} ≠ ∅ and computes its dimension. Under the Section 3 definition those intersections are empty; the arguments only work if U_G is the closure of the open cell. The paper never says that, and it never verifies that the incidence relations of these closures are preserved by the homeomorphism h (which would follow from h(cl(U_G)) = cl(h(U_G))). Proposition 4.10 depends on exactly that: it uses "nonempty intersection of codimension one" to decide which edge is added, and without the closed-cell reading the induction has no base. There is a second, smaller gap inside the converse of Proposition 4.8: the line "m=0 implies G⊂G'" needs an argument using endpoints of T(X); otherwise the codimension-one condition could in principle come from replacing one branch with another. The stress-test note is correct on both points.

These are repairable. The intended fix is to define V_G = cl(U_G) and do all of Section 4 with V_G, or at least state that U_G now means the closed cell. Add the missing lemma in Proposition 4.8, and the construction should go through. The theorem itself seems coherent, and the induction has the right shape. I do not see a circular or fitted assumption; the use of their own published Corollary 3.4 for order invariance is legitimate.

For hyperspace theorists working on uniqueness questions, this is a paper worth engaging with. A serious referee should see it, and should insist on the closed-cell clarification. If that is fixed, it is publishable; as written it is conditional.

Referee Report

3 major / 5 minor

Summary. The paper studies the pointed hyperspace C(p,X) of a tree X, consisting of all subcontinua containing p. It introduces a decomposition of the locally Euclidean part U(X) of C(p,X) into open cells indexed by subtrees of the 'core' T(X), and uses this cell structure to reconstruct the tree from the hyperspace. The main theorem (Theorem 4.14) asserts that for any tree X and p in X, the pair (X,p) is unique in the class of trees: if C(p,X) is homeomorphic to C(q,Y) for a tree Y and q in Y, then there is a homeomorphism X -> Y mapping p to q. The proof first treats arcs and n-ods (Theorems 4.3 and 4.4), then for the remaining trees uses the cell structure to recover the vertex set and edge relations of T(X) (Theorem 4.11), and finally extends from ramification points to ordinary and end points (Theorem 4.13).

Significance. Assuming the main theorem, this is a valuable contribution: it gives a uniqueness theorem for a pointed hyperspace in the class of trees, complements existing uniqueness results for hyperspaces C(X), and yields a partial solution to a problem on the size of K(X) (Corollary 4.15). The examples in Section 1 provide useful context, showing that the tree class is in some sense optimal, and the cell-decomposition method is natural and potentially adaptable to dendrites (as posed in Question 4.16). However, the proof as written has serious flaws in the cell-structure part, so the main theorem is not established.

major comments (3)

[Section 3 / Section 4 (Prop. 3.1(c), Props. 4.7, 4.8, 4.10)] The notation U_Y is introduced in Section 3 as an open subset of C(p,X), and Proposition 3.1(c) shows that the U_Y for different subtrees Y are pairwise disjoint open sets. Section 4, however, uses the same symbol U_G and asserts in Proposition 4.7 that U_G ∩ U_G' is nonempty for subtrees with certain incidence relations, and in Proposition 4.8 that for G' = G ∪ e the intersection is nonempty with dimension dim U_G - 1. These assertions are false for the open sets defined in Section 3: an element of U_G contains only a proper initial segment of an incident edge e, whereas an element of U_{G ∪ e} must contain the full edge e, so the two open sets are disjoint. The proofs need to work with the closures V_G = cl(U_G), or with an explicit cell-complex structure on C(p,X), and to show that the face-incidence relations of these closed cells are preserved by homeomorphisms. This is not a cosmetic issue: the base case of the induction in Proposition 4.10 relies on the asserted nonempty intersection U_G ∩ U_{G ∪ e}, which is empty under the Section 3 definition.
[Proposition 4.8, converse] The converse direction of Proposition 4.8 is not proved and is in fact false even under the closed-cell interpretation. From the condition dim(U_G ∩ U_G') = dim(U_G) - 1 the authors deduce l' + m = 1 and then state 'm = 0 implies that G ⊂ G'.' This implication is unjustified: m counts only edges outside G ∪ G' incident on G but not on G', while edges of G \ G' (the quantity l) are not constrained by the equation l' + m = 1. Concretely, let X be a tree with a ramification point p1 and three internal edges p0p1, p1p2, p1p3, with p0, p2, p3 not endpoints, and take G = {p0p1} ∪ {p1p2} and G' = {p0p1} ∪ {p1p3}. Then n = 0, l = l' = 1, m = m' = 0, so dim U_G = 1, dim U_G' = 1, and dim(U_G ∩ U_G') = 0 = dim U_G - 1, yet G' is not G ∪ e. Thus the converse cannot serve as the engine for concluding in Proposition 4.10 that h(U_{G ∪ e}) = U_{F ∪ f'}; the reconstruction of the path may fail.
[Proposition 4.10] The proof of Proposition 4.10 mixes the open-cell and closed-cell interpretations of U_G. For nested pairs G ⊂ G' it requires U_G ∩ U_G' ≠ ∅ (the closed-cell reading), while for a proper prefix G_i of the path and the extended path G' it asserts U_{G_i} ∩ U_{G'} = ∅ (the open-cell reading, since distinct components are disjoint by Proposition 3.1(c)). Under the closed-cell reading the latter disjointness is false, since the closure of U_{G_i} contains the closure of U_{G'} whenever G_i ⊂ G'. Under the open-cell reading the former nonempty intersection is false, as noted above. The same symbol U_G therefore cannot support both claims, and the induction step in Proposition 4.10 is internally inconsistent.

minor comments (5)

[Section 2] There is a typo: 'clousure' should be 'closure'.
[Theorem 4.4] The phrase 'Y is an simple ord(x,X)-od' should be 'Y is a simple ...-od'.
[Lemma 4.5] The induction step in the proof of Lemma 4.5 contains a confusing notational slip: the line 'Let ek+1 = f1, . . . , fm be the edges incident on G' appears to mix the labels of the newly added edge and the incident-edge set; this should be rewritten for clarity.
[Theorem 4.13] The reduction from the endpoint case to the ordinary-point case by attaching arcs is only sketched; a few more details would help the reader verify that the resulting homeomorphism of hyperspaces really forces the original trees to be homeomorphic.
[Corollary 3.4] The equality U_{ord(p,X)}(X) = U_p is stated without proof; since it requires that the component U_p is the unique component of dimension ord(p,X), a brief justification would improve the exposition.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the uniqueness theorem is proved from a homeomorphism-invariant cell decomposition of C(p,X), not assumed as input.

full rationale

The proof of Theorem 4.14 does not assume the uniqueness conclusion. Section 3 establishes a bijection between subtrees of T(X) and components of U(X) (Corollary 3.2) and expresses their dimensions in terms of incident edges (Corollary 3.4). Section 4 shows that the incidence and codimension-one relations among the components are preserved by homeomorphisms (Propositions 4.7 and 4.8) and uses these relations inductively to recover the graph structure of T(X) (Proposition 4.10), then extends the resulting isomorphism from T(X) to all of X (Theorems 4.11 and 4.13). No fitted parameter is renamed as a prediction, and no definition makes the target statement true by construction. The citations to [1] and [13] are external published results (order invariance and descriptions of C(p,X) for arcs and n-ods) whose assumptions do not include the target theorem, so they are independent support rather than circular premises. A possible gap about whether the U_G used in Propositions 4.7 and 4.8 are the open components of Section 3 or their closures is a correctness concern, not a circularity, and therefore does not affect this verdict.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No free parameters or invented entities appear; this is a pure topology proof. The main inputs from prior literature are Pellicer's descriptions of C(p,X) for arcs, and the authors' earlier order-invariance result. The proof's internal constructions (T(X), Sub_p(T(X)), U_G) are defined within the paper and are not pulled in as fitted quantities.

assumptions (4)

domain assumption Order of a point in a graph is invariant under homeomorphism of C(p,X) (Prop 4.1, cited to [1, Cor 3.4]).
Used as the first step to match p to q; taken from the authors' prior published paper.
domain assumption For an arc X, C(p,X) is an arc if p is an endpoint and a 2-cell otherwise (Pellicer [13, Thm 3.17]).
Used in Theorem 4.3 and Example 1.1 to identify arcs among trees.
domain assumption Hereditarily indecomposable continua have C(q,Y) an arc, giving non-uniqueness examples (Pellicer [13, Lemma 3.19]).
Used in Example 1.1 to show the general uniqueness statement is false.
standard math Standard facts about the Vietoris topology and the Hausdorff metric, including that C(p,X) is a continuum (Nadler [11]).
Background for all hyperspace arguments.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Uniqueness of the hyperspaces $C(p,X)$ in the class of trees." pith.science (2026). https://pith.science/paper/IYDDMRG6

@misc{pith2026190806202,
  author       = {Pith},
  title        = {Pith review of: Uniqueness of the hyperspaces $C(p,X)$ in the class of trees},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IYDDMRG6}},
  note         = {Machine review of arXiv:1908.06202}
}

abstract

Given a continuum $X$ and $p\in X$, we will consider the hyperspace $C(p,X)$ of all subcontinua of $X$ containing $p$. Given a family of continua $\mathcal{C}$, a continuum $X\in\mathcal{C}$ and $p\in X$, we say that $(X,p)$ has unique hyperspace $C(p,X)$ relative to $\mathcal{C}$ if for each $Y\in\mathcal{C}$ and $q\in Y$ such that $C(p,X)$ and $C(q,Y)$ are homeomorphic, then there is an homeomorphism between $X$ and $Y$ sending $p$ to $q$. In this paper we study some topological and geometric properties about the structure of $C(p,X)$ when $X$ is a tree, being the main result that $(X,p)$ has unique hyperspace $C(p,X)$ relative to the class of trees.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

16 extracted references · 11 canonical work pages

[1]

Corona–V´ azquez, R

F. Corona–V´ azquez, R. A. Qui˜ nones–Estrella, J. S´ anchez–Mart ´ ınez and H. Villanueva, Hyperspaces C(p, X ) of ﬁnite graphs , Topology Appl., 248 (2018), 40–49

2018
[2]

Eberhart, Intervals of continua which are Hilbert cubes , Proc

C. Eberhart, Intervals of continua which are Hilbert cubes , Proc. Amer. Math. Soc., 68 (2) (1978), 220–224

1978
[3]

Herrera–Carrasco, Dendrites with unique hyperspace , Houston J

D. Herrera–Carrasco, Dendrites with unique hyperspace , Houston J. Math., 33 (2007), 795–805

work page 2007
[4]

Herrera–Carrasco, M

D. Herrera–Carrasco, M. de J. L´ opez and F. Mac ´ ıas–Romero, Dendrites with unique symmetric products , Topology Proc., 34 (2009), 175–190

work page 2009
[5]

Illanes, The work of Sam B

A. Illanes, The work of Sam B. Nadler, Jr. on hyperspaces , Continuum Theory. Lecture Notes in Pure Appl. Math., 230, Marcel Dekker, Inc., New York and Basel, (2002) 9–31

work page 2002
[6]

Illanes, Finite graphs have unique hyperspaces Cn(X), Topology Proc., 27 (2003), 179–188

A. Illanes, Finite graphs have unique hyperspaces Cn(X), Topology Proc., 27 (2003), 179–188

work page 2003
[7]

Illanes, Uniqueness of hyperspaces , Questions Answers Gen

A. Illanes, Uniqueness of hyperspaces , Questions Answers Gen. Topol- ogy, 30 (2012), 21–44

work page 2012
[8]

Kuratowski, Topology Vol

K. Kuratowski, Topology Vol. II, New York: Acad. Press, 1968

work page 1968

Show all 16 references

[9]

Mac ´ ıas,On C-determined continua , Glas

S. Mac ´ ıas,On C-determined continua , Glas. Mat. Ser. III, 32 (52) (1997), 259–262

1997
[10]

J. M. Mart ´ ınez–Montejano, Models for Cp(X) for atriodic continua , Topology Appl., 154 (2007), 115–123

2007
[11]

S. B. Nadler, Jr., Hyperspaces of Sets: A Text with Research Ques- tions, Monographs and Textbooks in Pure and Applied Mathematics, 49, Marcel Dekker, Inc., New York and Basel, 1978

1978
[12]

S. B. Nadler, Jr., Continuum Theory: An introduction , Monographs and Textbooks in Pure and Applied Mathematics, 158, Marcel Dekker, Inc., New York and Basel, 1992. 16

1992
[13]

Pellicer, The hyperspaces C(p, X ), Topology Proc., 27 (1) (2003), 259–285

P. Pellicer, The hyperspaces C(p, X ), Topology Proc., 27 (1) (2003), 259–285

2003
[14]

Pellicer–Covarrubias, The hyperspaces K(X), Rocky Mountain J

P. Pellicer–Covarrubias, The hyperspaces K(X), Rocky Mountain J. Math., 35 (2) (2005), 655–674

2005
[15]

Pellicer–Covarrubias, The hyperspaces C(p, X ) for atriodic continua , Houston J

P. Pellicer–Covarrubias, The hyperspaces C(p, X ) for atriodic continua , Houston J. Math., 31 (2) (2005) 403–426

2005
[16]

Pellicer–Covarrubias, Cells in hyperspaces , Topology Appl., 154 (2007), 1002–1007

P. Pellicer–Covarrubias, Cells in hyperspaces , Topology Appl., 154 (2007), 1002–1007. 17

2007

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Corona–V´ azquez, R

F. Corona–V´ azquez, R. A. Qui˜ nones–Estrella, J. S´ anchez–Mart ´ ınez and H. Villanueva, Hyperspaces C(p, X ) of ﬁnite graphs , Topology Appl., 248 (2018), 40–49

2018

[2] [2]

Eberhart, Intervals of continua which are Hilbert cubes , Proc

C. Eberhart, Intervals of continua which are Hilbert cubes , Proc. Amer. Math. Soc., 68 (2) (1978), 220–224

1978

[3] [3]

Herrera–Carrasco, Dendrites with unique hyperspace , Houston J

D. Herrera–Carrasco, Dendrites with unique hyperspace , Houston J. Math., 33 (2007), 795–805

work page 2007

[4] [4]

Herrera–Carrasco, M

D. Herrera–Carrasco, M. de J. L´ opez and F. Mac ´ ıas–Romero, Dendrites with unique symmetric products , Topology Proc., 34 (2009), 175–190

work page 2009

[5] [5]

Illanes, The work of Sam B

A. Illanes, The work of Sam B. Nadler, Jr. on hyperspaces , Continuum Theory. Lecture Notes in Pure Appl. Math., 230, Marcel Dekker, Inc., New York and Basel, (2002) 9–31

work page 2002

[6] [6]

Illanes, Finite graphs have unique hyperspaces Cn(X), Topology Proc., 27 (2003), 179–188

A. Illanes, Finite graphs have unique hyperspaces Cn(X), Topology Proc., 27 (2003), 179–188

work page 2003

[7] [7]

Illanes, Uniqueness of hyperspaces , Questions Answers Gen

A. Illanes, Uniqueness of hyperspaces , Questions Answers Gen. Topol- ogy, 30 (2012), 21–44

work page 2012

[8] [8]

Kuratowski, Topology Vol

K. Kuratowski, Topology Vol. II, New York: Acad. Press, 1968

work page 1968