WorksMathematics
Property A and the identification of the maximal and reduced uniform Roe algebras
Abstract
Let be a discrete metric space of bounded geometry. We prove that has property A if and only if the canonical map is an isomorphism.
Introduction
Roe algebras arose in J. Roe’s study of higher indices of elliptic operators on open manifolds [11]. The reduced uniform Roe algebra appears in [12 Section 4.4], and its maximal counterpart follows the construction of the maximal Roe algebra by Gong, Wang and Yu [8 Sections 3–4].
For a discrete metric space of bounded geometry, the identity on the algebraic uniform Roe algebra extends to the canonical quotient map
Whether is an isomorphism is related to the large-scale geometry of . Yu introduced property A as a coarse analogue of amenability [17]. Špakula and Willett proved that property A implies that is an isomorphism and asked whether the converse holds [15 Proposition 1.3 and Section 1]. They also showed that induces an isomorphism on -theory when coarsely embeds into Hilbert space [15 Theorem 1.4]. The box space of constructed by Arzhantseva, Guentner and Špakula admits such an embedding and fails property A [3]. For that space, Špakula and Willett explicitly asked whether itself is an isomorphism [15 Example 1.16].
Question 1.1. Let be a discrete metric space with bounded geometry. If the canonical quotient map is an isomorphism, must have Yu’s property A?
Skandalis, Tu and Yu associated the coarse groupoid to and recast the coarse assembly map as a groupoid Baum–Connes assembly map with coefficients [14 Proposition 4.8]. The maximal and reduced uniform Roe algebras identify with the corresponding groupoid -algebras [15 Remark 1.5]
Under these identifications, is the canonical groupoid quotient. Skandalis, Tu and Yu also showed that has property A if and only if is topologically amenable [14 Theorem 5.3]. The question thus asks whether weak containment forces topological amenability for coarse groupoids.
Willett’s 2015 paper grew out of an attempt to answer this question for the box space of in Špakula and Willett’s Example 1.15 [16 Remark 3.5]. He constructed a Higson–Lafforgue–Skandalis groupoid with finite quotients of as its finite fibers and as the fiber at infinity. The non-amenability of the fiber at infinity forces the groupoid to fail topological amenability [16 Lemma 2.4]. Using Lubotzky and Shalom’s property FD, Willett showed that the regular representations of the chosen finite quotients recover the full norm on . This yields equality of the maximal and reduced groupoid -algebras [16 Lemmas 2.7 and 2.8]. In Remark 3.5 of the same paper, he asked whether weak containment can also hold for a non-amenable coarse groupoid.
Alekseev and Finn–Sell subsequently constructed non-amenable principal étale groupoids with weak containment and second-countable unit spaces [1 Section 2]. The coarse groupoid of an infinite has the non-second-countable Stone–Čech compactification as its unit space [14 Section 3]. Ara, Li, Lledó and Wu recorded the coarse groupoid question as open [2 Example 4.6]. For coarse disjoint unions of large-girth expanders, Buss, Echterhoff and Willett distinguished the two uniform Roe completions by comparing the spectra of the graph Laplacian [5 proof of Lemma 4.7]. We answer Question 1.1 affirmatively. Together with Špakula and Willett’s result [15 Proposition 1.3], this gives the following main theorem.
Theorem 1.2. Let be a discrete metric space with bounded geometry. Then, has Yu’s property A if and only if the canonical quotient map
is an isomorphism.
The proof was inspired by the work of Roe–Willett [13], in particular about the construction of the sequence of graphs using the failure off property A.
When fails property A, we use Elek’s result on uniform local amenability [6] to construct locally expanding graphs. We then construct large-girth graphs on subsets of their vertex sets, sign the edges using the theorem of Marcus, Spielman and Srivastava [10], and form the signed Laplacian . This gives an operator with
Together with Špakula and Willett’s forward implication [15 Proposition 1.3], this yields our main result.
This paper is organized as follows. In Section 2, we briefly recall the two uniform Roe completions, Yu’s property A, and the decomposition of finite-propagation operators into weighted partial translations. In § 3, we introduce the graph terminology and constructs large-girth graphs from local expansion. § 4 constructs a state on the maximal completion from signed paths. Section 5 establishes the norm estimates and proves Theorems 1.2.
AI disclosure. We thank ChatGPT-6 for assistance in understanding the result of Marcus, Spielman and Srivastava [10] and for language editing during the preparation of this manuscript. All main results and mathematical proofs were developed by the authors.
Uniform Roe algebras
In this section, we first briefly recall the definitions of maximal and reduced uniform Roe algebras and the concept of Yu’s property A.
Uniform Roe algebras
Let be a discrete metric space with bounded geometry. We write and set
Let . A bounded linear operator can be expressed as an -matrix
where for any . The propagation of is defined as
We say that the operator has finite propagation if
Now let us recall the definitions of the algebraic uniform Roe algebras.
Definition 2.1. Let be a discrete metric space with bounded geometry. The algebraic uniform Roe algebra, denoted by , is the -algebra of all bounded operators with finite propagation.
For any two operators , , the product is defined as
for any .
For every , there exists a constant such that
for any -representation and for any with $\operatorname{prop}(T)\le r [15, Lemma 1.10]. As a result, the following maximal norm is finite:
The maximal and reduced uniform Roe algebras can be viewed as coarse analogues of the maximal and reduced group -algebras.
Definition 2.2 ([15 Definitions 1.9 and 1.11]). Let be a discrete metric space with bounded geometry.
The uniform Roe algebra is defined to be the completion of under the norm on ;
The maximal uniform Roe algebra is defined to be the completion of under the maximal norm.
For the index-theoretic origins of Roe algebras, see [11]. In Roe’s terminology, is the uniform translation -algebra; see [12 Section 4.4].
By the universality of the maximal norm, the identity map on continuously extends to the quotient map
We use for the concrete operator norm.
We use the following equivalent formulation of Yu’s property A [17].
Definition 2.3 ([15 Definition 2.1]). Let be a discrete metric space with bounded geometry, and let denote the nonnegative real-valued functions in . We say that has Yu’s property A if for every and , there exists a map satisfying
for each , ;
there exists an , such that , for all ;
for all with , .
The following result of Špakula and Willett gives the forward implication of Theorem 1.2.
Theorem 2.4 ([15 Proposition 1.3]). Let be a discrete metric space with bounded geometry. If has Yu’s property A, then the quotient map
is an isomorphism.
Partial translations
A partial translation is a bijection between subsets of with . Its operator is
For , write for diagonal multiplication by .
Every finite-propagation operator is a finite sum of operators . Indeed, if the propagation is at most , each row and column contains at most nonzero entries. Enumerate the nonzero entries and color them greedily with colors, forbidding repetitions of a color in the same row or column. At most colors are forbidden at each step. Within each color, the nonzero entries define a partial translation from column indices to row indices, and their values define the coefficient on its range.
Local expansion and large-girth graphs
Let be a discrete metric space without property A. We first construct a sequence of finite graphs with uniform local expansion. Based on this sequence, we construct a sequence of graphs with degrees five or six, uniformly bounded ambient edge lengths, and girth tending to infinity.
Preliminaries on graphs
In this subsection, we recall some background on graph theory.
A simple graph is a pair , where is its vertex set and is a set of unordered pairs of distinct vertices. We write for the edge with endpoints and . The expressions and denote the same undirected edge. The endpoints of an edge are adjacent, or neighbors. The neighbor set and degree of are
An edge is incident to each of its endpoints. A graph has bounded degree if there is a finite constant such that for every vertex .
A subgraph of is a graph whose vertex set and edge set are subsets of and , respectively, and every edge of has endpoints in . For , the induced subgraph has vertex set and all edges of with both endpoints in . A simple graph on has vertex set , with an edge set specified separately.
A matching is a set of edges no two of which share an endpoint. A proper edge coloring assigns colors to edges so that edges sharing an endpoint have different colors. Each color class is therefore a matching.
A walk of length is a finite sequence
where has endpoints and . A walk is closed if . A trail is a walk with no repeated edge, and a path is a walk with pairwise distinct vertices. A graph is connected if any two vertices can be joined by a path. Its components are the maximal connected subgraphs. A cycle is a closed trail of positive length in which are pairwise distinct. Cycles are identified under change of starting vertex and reversal. Every closed trail of positive length contains a cycle, by taking a shortest nonempty closed subtrail.
The girth is the least length of a cycle, with value if there is no cycle. A forest is a graph with no cycles. A tree is a connected forest. In a tree, there is a unique path between any two vertices, since distinct paths with the same endpoints would contain a cycle in their union.
For vertices in the same component, the graph distance is defined to be the minimum length of a path joining them. For vertices in different components, we set .
If , the ambient length of the edge is . A bound on all ambient edge lengths implies
when are in the same component.
For nonempty , we write .
Definition 3.1. Let be a discrete metric space with bounded geometry and . The scale- graph has vertex set and an edge for each pair of distinct vertices with .
Definition 3.2. Let be a graph and . The internal vertex boundary and edge boundary are defined as
Every internal boundary vertex is incident to a crossing edge, so . If is a subgraph, counts only neighbors in and edges belonging to .
Local expansion
Brodzki, Niblo, Špakula, Willett and Wright introduced uniform local amenability (ULA) and proved its coarse invariance [4 Definition 2.2 and Lemma 2.6]. We use Elek’s formulation [6, 7] for graphs with bounded degree.
Definition 3.3 ([6 Definition 1.2]). Let be an infinite graph of bounded degree. The graph is said to be uniformly locally amenable (ULA) if, for every , there is such that each nonempty finite subgraph contains a nonempty with
For metric spaces with bounded geometry, Brodzki, Niblo, Špakula, Willett and Wright showed that Yu’s property A implies ULA [4 Proposition 3.2 and Lemma 2.7], while Elek proved the converse [6 Theorem 1]. For the equivalence in the graph-metric formulation, see also [7 Theorem 1].
The following construction is inspired by the arguments in Roe–Willett [13]. Failure of ULA gives rise to a sequence of finite graphs with a uniform lower bound on the ratio of internal boundary size to cardinality for small vertex sets.
Lemma 3.4. Suppose has bounded geometry without Yu’s property A. Then, there are constants , and an integer , and a sequence of finite connected simple graphs with nonempty, pairwise disjoint vertex sets in , such that
each has degrees at most and ambient edge lengths at most ;
for every and every nonempty subset with , we have that
Proof. It is obvious that the space is infinite. Suppose that every scale graph has ULA. Fix . Since has bounded degree, ULA gives nonnegative unit vectors supported in graph balls of a common radius , with on edges. Each such ball about is contained in , and distinct points at ambient distance at most are adjacent in . Thus gives the required unit vectors at scale , since . As and are arbitrary, has property A, a contradiction. Therefore, there must be an , such that the scaled graph does not have ULA.
Fix for which does not have ULA. There is such that for every positive integer there is a nonempty finite subgraph of satisfying
In particular, , since has empty internal boundary in . Let
where
We construct by induction. We start with . Choose satisfying (2) with and take a connected component with vertex set . A nonempty subset of ambient diameter at most lies in for any , so . Its internal boundary is the same in and , so (2) gives the required bound with constant .
For the construction of , take and . Choose an integer and a new satisfying (2) with . First remove . Let be the set of vertices removed so far and write for the remaining induced subgraph. Whenever contains a nonempty subset with
add to and repeat.
To see why this leaves a nonempty graph, count the internal boundary of in the original graph . When is added, a vertex in can remain on the boundary only if it was already there. The vertices of on the new boundary are exactly those with a neighbor outside , namely the vertices in . Thus each deletion adds at most boundary vertices. The initial boundary has size at most , and the sets removed at successive steps are disjoint. follows that
Initially . If first reaches or exceeds , then , since the last step adds at most vertices. Thus (2) applies to . But , so
contrary to (2). Hence throughout. The process terminates since is finite and each step removes at least one vertex. Since , the final is nonempty.
Take a connected component of the final and put . The initial deletion gives . If has ambient diameter at most , then for any , so . The stopping condition gives . Every neighbor in of a vertex of lies in , so .
Now suppose that have been constructed. Repeat the construction of with and . The resulting component avoids and satisfies the required boundary estimate with the same constant . Each is a subgraph of , so the degree bound and the ambient edge length bound hold uniformly in .
Definition 3.5. For a finite simple graph , the adjacency matrix acts on and is defined by
The degree matrix is , and the combinatorial Laplacian is .
With each undirected edge counted once, its quadratic form is
Let be the graphs with vertex sets in Lemma 3.4. On , consider the following operators
Each is symmetric and positive semidefinite, with nonnegative entries and row sums one. Consider the constant
For the operator , we have the following estimates on the diagonals of its powers , for .
Lemma 3.6. For and ,
Proof. Fix and . Let and let be the compression of to . For any function , it extends by zero to . List the distinct values of as , put , and set
Since , by the Lemma 3.4, we have that . Summing these inequalities with weights gives rise to the inequality
By Cauchy–Schwarz and the degree bound, the last sum is at most
It follows that , and hence
Thus .
Each multiplication by enlarges the support by at most in the ambient metric. Hence is supported in for . Projecting onto after each step therefore leaves these vectors unchanged, so . Accordingly, we have that
We use the estimates on the entries of to define a sequence of new graph and then removing certain vertices and edges gives us a sequence of graphs with large girth and degree at most 6.
Lemma 3.7. Let be as in Lemma 3.4, with vertex sets . There exist a constant , an integer and a sequence of simple graphs with nonempty vertex sets such that for each and
Proof. Fix an integer with . For , denote
The matrix is symmetric and positive semidefinite, with nonnegative entries and row sums one. By Lemma 3.6,
Let be the simple graph on with edge set
Since has propagation at most , every edge of has ambient length at most .
For a graph , write for its cycles of length less than . For , define
Set , so that . By symmetry and the row sums of ,
Choose a starting vertex and a direction for each -cycle of . Distinct cycles then give distinct terms in the expansion of . Each such term is the product of the corresponding edge weights divided by . Since the trace expansion has nonnegative terms and for , we obtain that
Consequently,
With the other coordinates fixed, is affine in . Thus one of the choices and preserves or increases . Applying this successively to all coordinates of gives with . Let be the graph on with edges . Then
For each choose , and let be obtained by deleting the set . Hence
Repeatedly delete vertices of current degree at most four from . If all vertices were deleted in the order , counting each edge when its first endpoint is deleted would result in
and this is a contradiction. The remaining graph therefore has a nonempty vertex set and satisfies
Starting from , delete edges whose two endpoints both have current degree greater than five until no such edge remains. Write for the resulting graph and put
Then is independent and for . For each , partition of its incident edges into pairs, where
Each pair forms a length-two path with distinct endpoints in . The chosen paths are pairwise edge-disjoint, since the pairs at each are disjoint and each edge of meets at most one vertex of .
Let consist of the chosen length-two paths together with the length-one paths given by the unused edges of . For a path with endpoints , write for its length and set . Define
If distinct had the same endpoints, traversing and returning along would give a closed trail in . Since a nonempty closed trail contains a cycle, this would give
contrary to . Thus the endpoint map is injective and is simple.
For , let be the number of chosen length-two paths with endpoint . Each such path replaces one edge incident to by one shortcut. For , each of the chosen paths removes two edges incident to . Hence
Let be the unique path representing . For a cycle in , concatenating the paths in cycle order gives a closed trail in , since the paths are pairwise edge-disjoint. Therefore
Finally, if , the triangle inequality along gives
The conclusion follows with and .
We would like to remark here that the shortcut construction follows from the arguments in [9 Lemma 14].
A state on the maximal uniform Roe algebra
In this section, we construct a state on the maximal uniform Roe algebra using products of edge signs along short paths, for any given signing of graphs. The state vanishes on the signed Laplacian and provides the lower bound for the maximal norm in § 5.
Let satisfy the conclusion of Lemma 3.7, with pairwise disjoint vertex sets . The degrees are five or six, the ambient edge lengths are at most , and . After discarding a finite initial segment, we have for every remaining . Set and for this sequence.
A signing of a graph is a map
Its signed adjacency matrix has entry when is an edge and zero otherwise. For a sequence of signings , let and be the block diagonal operators with blocks and the signed adjacency matrices, respectively, extended by zero off . The degree diagonal has norm at most six. For , Cauchy–Schwarz bounds by six times the sum of over neighbors of . Summing over counts each at most six times, so . The ambient edge length bound gives propagation at most . Define
The sign of a path is the product of its edge signs, with value one on the empty path. Since , choose integers with and . Write . For , let , and define
Two distinct paths of length at most with the same endpoints would contain a cycle of length at most in their union. The girth bound therefore makes the sign in the first case well defined. Define
Then is real and symmetric, its entries have absolute value at most one, , and on each edge of .
Lemma 4.1. Let be a positive integer. For any with , there is a positive semidefinite matrix on that differs from entrywise by at most .
Proof. Let be the set of indices in the average defining for which no pair satisfies . Since , the intervals are disjoint. Each unordered pair in therefore excludes at most one index, so at most indices lie outside .
Fix . If and , then by the triangle inequality. By the definition of , we have that . Thus the relation on is an
equivalence relation. Therefore, for vertices in distinct classes.
Fix a root in an equivalence class and write for the sign of the shortest path from to . For , the three shortest paths between have a connected union with at most edges, so their union is a tree. The signs on the common initial segment of the paths from to and from to cancel in , leaving the sign along the path from to . Hence on . Each block is therefore positive semidefinite, and so is .
Now set
Obviously, is positive semidefinite. Since each entry of has absolute value at most one, for we have
where the sum is over the indices in the average defining .
Now, we are ready to define the state using ultrafilter. A free ultrafilter on the infinite index set of is a family of subsets of closed under finite intersections and passage to supersets. It contains every cofinite subset of and exactly one of and for each . For a bounded sequence , its ultralimit
is defined by
The existence of this limit is guaranteed by the compactness of a closed disk containing the sequence, and it is obvious that the limit is unique.
Lemma 4.2. Fix a free ultrafilter on the indices. The formula
defines a functional on that extends uniquely to a state on satisfying
Proof. Fix with propagation at most . Each row and column has at most nonzero entries. At most terms can contribute to , and each has absolute value at most , since . Thus is bounded in for each , and the fixed ultralimit defines a linear functional . Since , we have for all , hence . Fix and denote
The row bound gives , while the column bound shows that only finitely many rows can meet . Expanding the matrix product gives rise to
For every intermediate row index , the set is contained in the ambient ball . Its cardinality is therefore bounded by , so Lemma 4.1 applies even when the vertices of are far apart in the graph metric. Replacing the kernel on by its positive semidefinite approximation makes the inner sum nonnegative. The original sum is real since is real and symmetric. The absolute value of the change is at most , where . By Cauchy–Schwarz inequality, we have that
Consequently,
For each fixed , the number is fixed while , so the error tends to zero. The ultralimit preserves inequalities between bounded real sequences, hence .
For the algebraic GNS construction, quotient by the null space of the positive form . If is a diagonal contraction or is a partial translation operator, then for a bounded diagonal operator . We may take in the first case and in the second. It follows that
Thus these left multiplications preserve the null space and act contractively on the quotient. By the decomposition in § 2.2, the left action of is bounded by . Left multiplication by is its adjoint for the form, since . Completing gives a bounded -representation . Since is a unit vector,
The representation extends to by the definition of the maximal norm, and its vector state at is the extension of .
The values of on the diagonal and on the edges determine the evaluations in (5) on and . On the diagonal , so is the average degree of . For each edge , the two ordered entries of contribute each. Hence
Taking the ultralimit gives and .
Proof of the main theorem
In this section, we shall prove our main theorem. Before proving, we shall recall the following theorem of Marcus–Spielman–Srivastava [10] to estimate the maximal norm of the operator of the signed adjacency matrices.
Theorem 5.1 ([10 Theorems 3.2, 3.7, 4.4 and 5.2]). Let be a nonempty finite simple graph with degree at most . Then, there is a signing
such that the largest eigenvalue of its signed adjacency matrix satisfies
For the graphs constructed in the Lemma 3.7, we take . The above eigenvalue bound gives an upper bound on in the concrete operator order. The minimum degree of vertices is five, and , so subtracting from the degree diagonal ensures a strictly positive lower bound for the signed Laplacian on .
Theorem 5.2. Let be a discrete metric space of bounded geometry. If does not have property A, then the quotient map is not an isomorphism.
Proof. Suppose that does not have property A. Let be the sequence of graphs obtained in Lemmas 3.4 and 3.7. By Theorem 5.1, for any a signing on each , the signed adjacency matrix has largest eigenvalue at most . Let
Since the disjoint union of the has maximum degree six, a greedy edge coloring uses at most eleven colors. Each color class is a matching. Orient its edges and let be the resulting partial translation operator weighted by the edge signs. Each belongs to because its displacement is at most . Writing , we have . By summing over the matchings, we have that
Therefore,
in . The diagonal has a square root in , so , and hence
In the reduced completion, the signing bound gives , while . It follows that
On the maximal completion, the state in Lemma 4.2 staisfies that
Thus, . It follows from that
Set by continuous functional calculus. Since is positive and has maximal norm one, its spectrum in contains , so . On the other hand, , and therefore
Thus is a nonzero element of . □
Together with [15 Proposition 1.3], we have the following:
Theorem 5.3. Let be a discrete metric space with bounded geometry. Then, has property A if and only if the canonical quotient map
is an isomorphism.
Jintao Deng
Department of mathematics, Center of Noncommutative Geometry, Shanghai University of Finance and Economics
Email: [email protected]
Geng Tian
Department of Mathematics, Liaoning University, Liaoning, China 110036
Email: [email protected]
References
- [1]V. Alekseev and M. Finn-Sell, Non-amenable principal groupoids with weak containment, Int. Math. Res. Not. IMRN (2018), 2332–2340.
- [2]P. Ara, K. Li, F. Lledó and J. Wu, Amenability and uniform Roe algebras, J. Math. Anal. Appl. 459 (2018), 686–716.DOI
- [3]G. Arzhantseva, E. Guentner and J. Špakula, Coarse non-amenability and coarse embed-dings, Geom. Funct. Anal. 22 (2012), 22–36.
- [4]J. Brodzki, G. A. Niblo, J. Špakula, R. Willett and N. Wright, Uniform local amenability, J. Noncommut. Geom. 7 (2013), 583–603.arxiv.org/abs/1203.6169
- [5]A. Buss, S. Echterhoff and R. Willett, The maximal injective crossed product, Ergodic Theory Dynam. Systems 40 (2020), 2995–3014.DOI
- [6]G. Elek, Uniform local amenability implies property A, Proc. Amer. Math. Soc. 149 (2021), 2573–2577.arxiv.org/abs/1912.00806
- [7]G. Elek and Á. Timár, Strong almost finiteness, J. Funct. Anal. 289 (2025), Article 111116.arxiv.org/abs/2308.14554
- [8]G. Gong, Q. Wang and G. Yu, Geometrization of the strong Novikov conjecture for residually finite groups, J. Reine Angew. Math. 621 (2008), 159–189.DOI
- [9]G. Kun, Expanders have a spanning Lipschitz subgraph with large girth, arXiv:1303.4982v2 (2013).arxiv.org/abs/1303.4982
- [10]A. W. Marcus, D. A. Spielman and N. Srivastava, Interlacing families I: Bipartite Ramanujan graphs of all degrees, Ann. of Math. 182 (2015), 307–325.arxiv.org/abs/1304.4132
- [11]J. Roe, An index theorem on open manifolds. I, J. Differential Geom. 27 (1988), 87–113.DOI
- [12]J. Roe, Lectures on coarse geometry, University Lecture Series, vol. 31, Amer. Math. Soc., Providence, RI, 2003.DOI
- [13]J. Roe and R. Willett, Ghostbusting and property A, J. Funct. Anal. 266 (2014), no. 3, 1674–1684.
- [14]G. Skandalis, J.-L. Tu and G. Yu, The coarse Baum–Connes conjecture and groupoids, Topology 41 (2002), 807–834.DOI
- [15]J. Špakula and R. Willett, Maximal and reduced Roe algebras of coarsely embeddable spaces, J. Reine Angew. Math. 678 (2013), 35–68.arxiv.org/abs/1110.1531
- [16]R. Willett, A non-amenable groupoid whose maximal and reduced C*-algebras are the same, Münster J. Math. 8 (2015), 241–252.arxiv.org/abs/1504.05615
- [17]G. Yu, The coarse Baum–Connes conjecture for spaces which admit a uniform embedding into Hilbert space, Invent. Math. 139 (2000), 201–240.DOI