text
stringlengths
128
2.05k
[EQUATION] since [EQUATION] and [MATH] Note that the proof above also shows that if [MATH] is the set of cancellations between [MATH] , then for any cancellation [MATH] of [MATH] we have
[EQUATION] Now we have a tool to bound the length of a cancellation between two words in [MATH] from above. The next lemma bounds the length of a word in [MATH] from below. We will in the following call [MATH]
positive if [MATH] consists only of positive powers of [MATH] . Similarly, we say [MATH] is negative if [MATH] consists only of negative powers of [MATH] . Note that if [MATH] is either positive or negative then we clearly have [MATH]
Lemma 8.9 Let [MATH] and [MATH] . Then [EQUATION] Proof. It is enough to consider [MATH] Moreover, it is clear that if [MATH] , then
[EQUATION] since there is no cancellation due to the fact that [MATH] is either positive or negative. For the remaining cases, we will begin with some observations. Assume [MATH] satisfy that [MATH] is reduced, [MATH] is positive and [MATH] is negative. Moreover, let [MATH] and [MATH] be such that [MATH] . Then we have...
Now let [MATH] be neither positive or negative and let [MATH] . Fix [MATH] and [MATH] such that [MATH] and [MATH] , for all [MATH]
Assume first that for all [MATH] we have [MATH] is neither positive nor negative. Then for each [MATH] , we may fix [MATH] together with positive [MATH] , and negative [MATH] , such that for each [MATH] there is a cyclically reduced cyclic conjugate of [MATH] and of [MATH] of the form
[EQUATION] and [EQUATION] respectively. Then [MATH] and hence, by the observations above, we have [EQUATION] Next, assume that [MATH] is least such that [MATH] is either positive or negative. Then, as before, we may for each [MATH] chose [MATH] together with positive [MATH] and negative [MATH] such that for each [MATH]...
[EQUATION] and [EQUATION] respectively. Then [MATH] and hence, by the observations above, we have [EQUATION] Finally, to finish the proof, note that for any [MATH] , we can chose [MATH] in the argument above. Thus, as
[EQUATION] we obtain [EQUATION] as wanted. Note that [EQUATION] since [MATH] . This will be used all the time below to conclude that there is no bad cancellation in the various cases.
We will now begin to argue that there is no bad cancellation between two words from [MATH] . So let [MATH] . The following decomposition will be useful. For [MATH] let
[EQUATION] [EQUATION] and for [MATH] let [EQUATION] Then for [MATH] we have, up to cyclic permutation, that [EQUATION] and [EQUATION]
is a reduced word whenever each factor is reduced, for all [MATH] . However, it is not necessarily cyclically reduced. If for some [MATH] we have [MATH] or [MATH] or [MATH] , then
[EQUATION] is cyclically reduced. If [MATH] and [MATH] , then any possible reduction in the induced cycle of [MATH] is contained in [MATH]
Claim 1: If [MATH] and [MATH] , then there is no bad cancellation between [MATH] and [MATH] Proof. It is easy to check that [MATH] satisfies the assumption of Lemma 8.8 , since [MATH] . Thus any cancellation, [MATH] , between [MATH] and [MATH] satisfies
[EQUATION] So, by Lemma 8.9 , there cannot be any bad cancellation between [MATH] and [MATH] In the following, we let [EQUATION]
Note that [MATH] Claim 2: If [MATH] with [MATH] and [MATH] , then there is no bad cancellation between [MATH] and [MATH] Proof. Note that either [MATH] or [MATH] . Assume without loss of generality that we are in the first case. Then the only powers of [MATH] occurring in [MATH] are [MATH] and [MATH] . Thus any cancell...
[EQUATION] Moreover, by Lemma 8.9 , it holds that [EQUATION] since [MATH] . Thus there is no bad cancellation between [MATH] and [MATH]
Now we will take care of the case where [MATH] are different, but none of them extends the other. Claim 3: If [MATH] with [MATH] [MATH]
[EQUATION] then there is no bad cancellation between [MATH] and [MATH] Proof. First note that [MATH] will only contain [MATH] as powers of [MATH] , while [MATH] will only contain [MATH] as powers of [MATH]
We now claim that for each [MATH] and [MATH] we have [MATH] contains the string [MATH] or [MATH] , for some [MATH] . Indeed [MATH] is of one of the forms [MATH] [MATH] [MATH] or [MATH] , for some [MATH] and [MATH] . By straightforward calculation the statement is clearly true for [MATH] or [MATH] . To see that the stat...
Similarly, for each [MATH] and [MATH] , we have [MATH] contains the string [MATH] or [MATH] for some [MATH] Now let [MATH] satisfy that [MATH] and [MATH] . Then from the above it follows that any cancellation, [MATH] , between [MATH] and [MATH] is contained in either
[EQUATION] [EQUATION] or in one of their inverses, for some [MATH] Therefore, by Lemma 8.8 , we have [EQUATION] So, by Lemma 8.9 , we have [MATH]
Similarly any cancellation, [MATH] , between [MATH] and [MATH] is contained in either [EQUATION] [EQUATION] or in one of their inverses, for some [MATH] . So, by Lemma 8.8 and Lemma 8.9 , we also have [MATH] . Thus there cannot be any bad cancellation between [MATH] and [MATH]
From Claim 1, Claim 2 and Claim 3 we may conclude that there is no bad cancellation between two words in [MATH] , i.e., that [MATH] satisfies the [MATH] cancellation property.
We will now prove that there is no bad cancellation between a word from [MATH] and a word from [MATH] . To do so, let [MATH] [MATH] and
[EQUATION] Then [MATH] . We will again consider fixed [MATH] . Let [MATH] be fixed such that [MATH] and [MATH] Claim 4: If [MATH] with [MATH] and [MATH] , then there is no bad cancellation between [MATH] and [MATH]
Proof. This follows by the same arguments as the ones used in the beginning of the proof of Claim 2. Claim 5: If [MATH] [MATH] or [MATH] [MATH] . Then there is no bad cancellation between [MATH] and [MATH]
Proof. This follows by arguments similar to those in the beginning of the proof of Claim 3. From Claim 4, we may assume that both [MATH] . Moreover, by Claim 5, there is no bad cancellation between [MATH] and [MATH] in the case [MATH] and [MATH] or in the case [MATH] and [MATH] . In the next three claims, we prove that...
Claim 6: If [MATH] , then there is no bad cancellation between [MATH] and [MATH] Proof. Consider [MATH] for [MATH] and [MATH] . Either all the powers of [MATH] are positive or all the powers of [MATH] are negative. Below we have put these observations into a table. Here [MATH] and [MATH] refer to the sign of the occurr...
[MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] It is easily seen that if the signs of the powers of [MATH] do not match, then there is no bad cancellation between [MATH] and [MATH] for the corresponding [MATH] . So assume that the sign of the powers of [MATH] in [MATH]...
[EQUATION] [EQUATION] or in one of their inverses, for some [MATH] . Similarly, [MATH] is also contained in either [EQUATION] [EQUATION]
or in one of their inverses for some [MATH] Thus, by 8.8 , we have [EQUATION] and hence, by 8.9 , there is no bad cancellation between [MATH] and [MATH]
Claim 7: If [MATH] , then there is no bad cancellation between [MATH] and [MATH] Proof. First, consider the sign of the powers of [MATH] occurring in [MATH] and [MATH] for [MATH]
[MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] [MATH] It is easily seen that if the sign in two cells does not match, then there is no bad cancellation between [MATH] and [MATH] , for the corresponding [MATH] In case the signs of the powers of [MATH] are the same, one may use...
Claim 8: If [MATH] , then there is no bad cancellation between [MATH] and [MATH] Proof. First, by considering the form of [MATH] and [MATH] for the words [MATH] , one finds that for all [MATH] , we have [MATH] and [MATH] will contain at most one occurrence of one of the strings
[EQUATION] for some [MATH] . Moreover, for all [MATH] and [MATH] , we have that [MATH] and [MATH] contain at least one of the strings
[EQUATION] for some [MATH] This is again straightforward to check, by considering the form of [MATH] and [MATH] Note that the above implies that [MATH] contains at most one of the strings
[EQUATION] for some [MATH] . Moreover, for all [MATH] , we have that [MATH] contains at least one of the strings [EQUATION] for some [MATH]
Conversely, the above also implies that [MATH] contains at most one of the strings [EQUATION] for some [MATH] . Moreover, for all [MATH] , we have that [MATH] contains at least one of the strings
[EQUATION] for some [MATH] Thus, by making considerations and use of 8.8 and 8.9 , as in the proof of the earlier claims, we may conclude that there cannot be any bad cancellation between [MATH] and [MATH]
Claim 9: If [MATH] and [MATH] , then there is no bad cancellation between [MATH] and [MATH] Proof. From earlier results it is enough to consider the case where [MATH] or [MATH] . Assume without loss of generality that we are in the first case. Then there exist [MATH] such that [MATH] and [MATH] , for some [MATH] . Then...
[EQUATION] [EQUATION] or in one of their inverses. Therefore, by 8.8 we obtain [EQUATION] and hence, by 8.9 , we have [MATH] If [MATH] , then by 8.9 and the fact that [MATH] , we also have
[EQUATION] So assume [MATH] . Then we have that each of [MATH] and [MATH] occurs at most once in [MATH] . But by the arguments in Claim 3, we have that for all [MATH] there is [MATH] such that [MATH] or [MATH] occur in [MATH] . Thus [MATH] or [MATH] occur in [MATH] for all [MATH]
Therefore [MATH] is contained in either [EQUATION] [EQUATION] or in one of their inverses, for some [MATH] . Thus we obtain [EQUATION]
So, by 8.9 , we have [MATH] , as well. We may therefore conclude that there is no bad cancellation between [MATH] and [MATH] Putting all the claims together we finally have a proof of the following theorem.
Theorem 8.10 Let [MATH] for some [MATH] . Then there is a transversal [MATH] for the left cosets in [MATH] such that the set [EQUATION]
satisfies the [MATH] cancellation property. Using this, and the criterion in 8.1 , it is now easy to construct continuum many non-atomic invariant random subgroups of [MATH] , which are [MATH] -invariant and weakly mixing with respect to the action of [MATH]
Fix by 8.10 a transversal [MATH] for the left cosets in [MATH] such that the set [MATH] satisfies the [MATH] cancellation property. Moreover, fix an enumeration [MATH] such that [MATH] is the identity. Then we have
[EQUATION] since it follows by 8.9 that [MATH] for all [MATH] . Thus, by Remark 8.2 , condition [MATH] in 8.1 is satisfied, and hence the proof of [MATH] in 8.1 provides a family as the one described above.
There is also another consequence of 8.10 Let [MATH] denote the set of conjugacy classes of [MATH] and let [MATH] be the outer automorphism group of [MATH] . Now consider the natural action [MATH] . It is well known that there is a conjugacy class [MATH] such that [MATH] (see LS77 , page 45] ). Therefore for such [MATH...
[EQUATION] With 8.10 we obtain the following strengthening of this result. Corollary 8.11 There exists a conjugacy class [MATH] such that
[EQUATION] that is, [MATH] is disjoint from the (normal) subgroup [MATH] for [MATH] Proof. Let [MATH] be the conjugacy class of [MATH] for some [MATH] and assume towards contradiction that
[EQUATION] Now choose by 8.10 a transversal [MATH] for the left cosets in [MATH] such that [MATH] satisfies the [MATH] cancellation property. Fix [MATH] such that [MATH] . Then we have
[EQUATION] which contradicts that [MATH] for all [MATH] We do not know if the analogs of 8.10 and 8.11 hold for any [MATH] with [MATH]
Department of Mathematics California Institute of Technology Pasadena, CA 91125 kechris@caltech.edu Department of Mathematical Sciences University of Copenhagen Universitetsparken 5 DK-2100 Copenhagen vibquo@math.ku.dk
# Source: arxiv 1806.08677 # Title: Bounded G-theory with fibred control # Sections: all # Downloaded: 2026-03-03T05:20:09.790540+00:00
Bounded -theory with fibred control Abstract. We use filtered modules over a Noetherian ring and fibred bounded control on homomorphisms to construct a new kind of controlled algebra with applications in geometric topology. The resulting theory can be thought of as a “pushout” of bounded [MATH] -theory with fibred cont...
1. Introduction The purpose of this paper is to use filtered modules over a Noetherian ring with a fibred bounded control on homomorphisms to construct a bounded [MATH] -theory with fibred control. This theory can be thought of as a “pushout” of the bounded [MATH] -theory with fibred control constructed by the authors ...
and the controlled [MATH] -theory constructed in Here is a summary of this situation: Throughout this paper, metric spaces such as [MATH] and [MATH] that appear in the square will be proper metric spaces in the sense that every closed bounded subspace is compact. The space or spectrum in the upper left corner represent...
, reviewed here in the beginning of section 2.1 . This theory is defined for any ring of coefficients [MATH] . It is built out of free [MATH] -modules with generating sets parametrized over the metric space [MATH] . This allows to impose geometric control conditions on the homomorphisms [MATH] . The bounded control con...
The spectrum in the upper right corner [MATH] is a generalization of this theory to the situation when the modules are parametrized by the product of two metric spaces [MATH] and [MATH] , and the control imposed on the homomorphisms is relaxed: it is essentially the bounded control across [MATH] but the bound is allowe...
To describe the bottom row in the square and for the rest of the paper, we restrict to Noetherian rings [MATH] In place of parametrizations used to control homomorphisms between free modules, one can use filtrations of arbitrary [MATH] -modules by subsets of the metric space [MATH] and impose control conditions in term...
for a single space [MATH] . The result was the bounded [MATH] -theory spectrum [MATH] . The definition involved promoting the setting from the additive structure for free modules in the definition of bounded [MATH] -theory to a specific non-split Quillen exact structure on a category of filtered [MATH] modules with mor...
Now it is clear what the “pushout” [MATH] is supposed to mean. We want to look at the [MATH] -theory of a category built out of filtered modules over the product [MATH] where the morphisms have the fibred control condition of the type described for fibred [MATH] -theory. This time we are interested in very specific exc...
Acknowledgement. We would like to thank the referee for excellent comments that improved the narrative and the precision of the paper.
2. Elements of bounded -theory Bounded [MATH] -theory defined in is a variant of bounded [MATH] -theory of Pedersen and Weibel made applicable to more general, non-split exact structures. It was designed by the authors for a different purpose than the one in this paper. The old focus was on the equivariant theory in ad...
in the form best fit for the fibred theory. 2.1. Basic definitions We start with a brief recollection of the bounded [MATH] -theory setup. The coefficients [MATH] for this theory can be an arbitrary associative ring. The bounded category
[MATH] is the additive category of geometric [MATH] -modules whose objects are functions [MATH] which are locally finite assignments of free finitely generated [MATH] -modules [MATH] to points [MATH] of [MATH] The local finiteness condition requires precisely that for any bounded subset [MATH] the restriction of [MATH]...
[EQUATION] with the property that the components [MATH] are zero for [MATH] for some fixed real number [MATH] The associated [MATH] -theory spectrum is denoted by [MATH]
and is called the bounded K-theory of [MATH] 2.1.1 Remark We would like to remind the reader that the original paper of Pedersen and Weibel
was already written in greater generality. If [MATH] is any additive category, it can be used as coefficients in this construction in place of finitely generated free [MATH] -modules using the same formulas as above. The outcome is the bounded category [MATH] which is again an additive category with the evident notion ...
Pedersen and Weibel were more interested in the product situation and, in fact, repaired the non-uniform boundedness properties of morphisms in [MATH] by filtering the morphism sets. With that fix the category becomes isomorphic to [MATH] . We, instead, embraced the flexibility of this construction in
with the idea of exploiting the additional deformations in [MATH] , the K-theory with fibred control , that the construction allows. This is the spectrum that shows up in Figure as [MATH]
2.1.2 Notation For a subset [MATH] and a real number [MATH] [MATH] will stand for the metric [MATH] -enlargement [MATH] In this notation, the metric ball of radius [MATH] centered at [MATH] is [MATH] or simply [MATH]
A variation of the basic construction of bounded [MATH] -theory is based on the following observation. For every object [MATH] and a subset [MATH] there is a free [MATH] -module [MATH] . In this context we say an element [MATH] is supported on a subset [MATH] if [MATH] Now the restriction from arbitrary [MATH] -linear ...
In the rest of this section and the rest of the paper, we will restrict to the case of a Noetherian ring [MATH] Let [MATH] denote the power set of [MATH] partially ordered by inclusion and viewed as a category. If [MATH] is a left [MATH] -module, let [MATH] denote the family of all [MATH] -submodules of [MATH] partiall...
2.1.3 Definition An [MATH] -filtered [MATH] -module is a module [MATH] together with a functor [MATH] from the power set of [MATH] to the family of [MATH] -submodules of [MATH] , both ordered by inclusion, such that the value on [MATH] is [MATH] It will be most convenient to think of [MATH] as the functor above and use...
reduced if [MATH] An [MATH] -homomorphism [MATH] of [MATH] -filtered modules is boundedly controlled if there is a fixed number [MATH] such that the image [MATH] is a submodule of [MATH]
for all subsets [MATH] of [MATH] The objects of the category [MATH] are the reduced [MATH] -filtered [MATH] -modules, and the morphisms are the boundedly controlled homomorphisms.
The category [MATH] we constructed is clearly an additive category, but the more interesting structure for developing its [MATH] -theory is a certain Quillen exact structure. For a good modern exposition of exact categories we refer to Keller
; there is also a leisurely review of relevant basic theory in , section 2] Let us recall some standard terms. If a category has kernels and cokernels for all morphisms, it is called preabelian . If, in addition, the canonical map [MATH] for each morphism [MATH] is monic and epic but not necessarily invertible, we will...
semi-abelian , cf. 16 , pages 167-168] and . It is abelian if it is also balanced in the sense that the canonical map is an isomorphism. Recall also that a category is called cocomplete if it contains colimits of arbitrary small diagrams, cf. Mac Lane 13 , chapter V]
2.1.4 Remark If [MATH] is unbounded, [MATH] is not a balanced category and therefore not an abelian category. For an explicit description of a boundedly controlled morphism in
[MATH] which is an isomorphism of left [MATH] -modules but whose inverse is not boundedly controlled, we refer to 14 , Example 1.5]
It turns out that the kernels and cokernels in [MATH] can be characterized using an additional property a boundedly controlled morphism may or may not have.
2.1.5 Definition A morphism [MATH] in [MATH] is called boundedly bicontrolled if there exists a number [MATH] such that in addition to inclusions of submodules
[EQUATION] there are inclusions [EQUATION] for all subsets [MATH] In this case we will say that [MATH] has filtration degree [MATH] and write [MATH]
2.1.6 Definition We define the admissible monomorphisms in [MATH] be the boundedly bicontrolled homomorphisms [MATH] such that the map [MATH] is a monomorphism. We define the admissible epimorphisms be the boundedly bicontrolled homomorphisms [MATH] such that [MATH] is an epimorphism.
Let the class [MATH] of exact sequences consist of the sequences [EQUATION] where [MATH] is an admissible monomorphism, [MATH] is an admissible epimorphism, and [MATH]
There are numerous examples of semi-abelian categories that are classical or have appeared recently in analysis and algebra that are listed in section 4 of
or in . For us [MATH] is of major interest. 2.1.7 Theorem [MATH] is a cocomplete semi-abelian category. The class of exact sequences [MATH] gives an exact structure on [MATH]
Proof. This fact is contained in Proposition 2.6 and Theorem 2.13 of We want to recall the explicit construction of kernels and cokernels [MATH] for future reference. For a boundedly controlled morphism
[MATH] , the kernel of [MATH] in [MATH] has the [MATH] -filtration [MATH] where [MATH] This gives a kernel of [MATH] in [MATH] Similarly, let [MATH] be the [MATH] -filtration of the image of
[MATH] in [MATH] by [MATH] . Let [MATH] be the cokernel of [MATH] in [MATH] , which is the quotient [MATH] Then there is a filtration [MATH] of [MATH] defined by [MATH] where the maps between quotients are induced from the structure maps of [MATH] The resulting epimorphism [MATH]
gives a boundedly bicontrolled morphism of filtration [MATH] . Lemma 2.5 of verifies the universal properties of the cokernel [MATH]
Now suppose [MATH] is an arbitrary cocomplete semi-abelian category. All of the definitions and proofs presented so far can be interpreted verbatim to describe a bounded category [MATH] if instead of [MATH] -modules one uses objects from [MATH] . We should point out that the exposition in
is constructed in lesser generality with [MATH] assumed to be abelian. It is important for the fibred version to relax abelian to semi-abelian. However,
can still serve as a good reference because throughout that paper only semi-abelian properties of [MATH] are used. In particular we have this conclusion.
2.1.8 Theorem For any cocomplete semi-abelian category [MATH] [MATH] is a cocomplete semi-abelian category. Of course, the category of [MATH] -modules [MATH] is a cocomplete abelian category and so can serve as a basic example of cocomplete semi-abelian coefficients [MATH] . In this case [MATH] is precisely [MATH]