betaFeedback

WorksMathematics

Higher-order derivatives and their applications to diffeomorphism groups

Abstract

In this paper, we apply higher-order tangent bundles and higher-order derivative maps to study countable subgroups of the diffeomorphism group Diff(M){\rm Diff}(M) of a closed smooth manifold MM. Using this framework, we introduce the concept of higher-order μμ-discreteness for subgroups Γ<Diff(M)Γ< {\rm Diff}(M), and show that both the strong Novikov conjecture and the coarse Baum–Connes conjecture hold for countable groups with higher-order μμ-discreteness. Finally, we give examples showing that higher-order μμ-discreteness encompasses a strictly larger class of subgroups of Diff⁡(M)\operatorname{Diff}(M) than its lower-order counterpart.

Introduction

The Novikov conjecture asserts that the higher signatures of closed oriented smooth manifolds are invariant under orientation-preserving homotopy equivalences. Noncommutative geometry provides an approach to the study of the Novikov conjecture via higher index theory. More precisely, for a countable discrete group Γ\Gamma, the Novikov conjecture for closed oriented smooth manifolds with fundamental group Γ\Gamma follows from the rational strong Novikov conjecture for Γ\Gamma. The latter asserts that the rational Kasparov–Baum–Connes assembly map

K∗(BΓ)⊗ZQ⟶K∗(Cr∗Γ)⊗ZQK_{*}(B\Gamma) \otimes_{\mathbb{Z}} \mathbb{Q} \longrightarrow K_{*}(C_{r}^{*}\Gamma) \otimes_{\mathbb{Z}} \mathbb{Q}

is injective, where K∗(BΓ)K_{*}(B\Gamma) denotes the KK-homology of the classifying space BΓB\Gamma for the group Γ\Gamma and K∗(Cr∗Γ)K_{*}(C_{r}^{*}\Gamma) denotes the KK-theory of the reduced group C∗C^{*}-algebra Cr∗ΓC_{r}^{*}\Gamma associated to Γ\Gamma. Although the Novikov conjecture has been established for many classes of

Key words and phrases. smooth diffeomorphisms, higher-order derivative, coarse embedding, Novikov conjecture.

The first author is partially supported by Shanghai Pujiang Project, No. 25PJA033.

The second author was supported in part by Grant for Excellent Young Scholars in Tianyuan Mathematics. groups, it remains open in general for a natural and important class: countable subgroups of diffeomorphism groups of compact smooth manifolds.

Inspired by Connes’ theorem [5] on higher signatures associated with Gelfand–Fuchs classes of diffeomorphism groups, Gong, Wu and Yu [8] introduced a notion of geometric discreteness to describe the extent of derivative distortion for the volume-preserving subgroups of Diff⁡(M)\operatorname{Diff}(M). They proved that every such geometrically discrete subgroup satisfies the rational strong Novikov conjecture. Gong, Wu, Xie and Yu [7] subsequently removed the volume-preserving assumption by introducing μ\mu-discreteness for subgroups of Diff⁡(M)\operatorname{Diff}(M) with respect to a Borel probability measure μ\mu on MM. They also proved the rational strong Novikov conjecture for every countable μ\mu-discrete subgroup of Diff⁡(M)\operatorname{Diff}(M).

Both geometric discreteness and μ\mu-discreteness are defined in terms of the first derivatives of diffeomorphisms.

In this paper, we use higher-order derivatives to study the diffeomorphisms. In local coordinates, the ordinary derivative records only the linear term of the Taylor expansion of a diffeomorphism. However, the rr-th order derivative encodes all partial derivatives of orders 1,…,r1,\ldots,r, thereby retaining the nonlinear terms through degree rr. This leads to the notion of higher-order μ\mu-discreteness. It turns out that lower-order μ\mu-discreteness implies higher-order μ\mu-discreteness, but the converse fails in general, as the example in § 4 shows. Thus, higher-order μ\mu-discreteness allows us to detect a strictly larger class of subgroups of Diff⁡(M)\operatorname{Diff}(M) than lower-order μ\mu-discreteness does.

Since our framework is based on coarse embeddings of the relevant spaces into LpL^p-spaces, it yields not only the strong Novikov conjecture but also the coarse Baum–Connes conjecture for all higher-order μ\mu-discrete groups.

Recall that a map h:(X,dX)→(Y,dY)h:(X,d_X)\to(Y,d_Y) between metric spaces is a coarse embedding [9] if there exist nondecreasing functions ρ−,ρ+:[0,∞)→[0,∞)\rho_-,\rho_+:[0,\infty)\to[0,\infty), with ρ−(t)→∞\rho_-(t)\to\infty as t→∞t\to\infty, such that

ρ−(dX(x,y))≤dY(h(x),h(y))≤ρ+(dX(x,y)),x,y∈X.\rho_-(d_X(x,y)) \le d_Y(h(x),h(y)) \le\rho_+(d_X(x,y)), \qquad x,y\in X.

The following is our main result.

Theorem 1.1. Let MM be a closed smooth manifold, and Γ<Diff⁡(M)\Gamma<\operatorname{Diff}(M) an (r,p)(r,p)-th order μ\mu-discrete subgroup. Then the strong Novikov conjecture holds for Γ\Gamma, i.e. the Baum–Connes assembly map

μ:KK∗Γ(EΓ)⟶K∗(C∗(Γ))\mu:KK_*^\Gamma(E\Gamma)\longrightarrow K_*(C^*(\Gamma))

is injective, where EΓE\Gamma is the universal space for proper actions of the group Γ\Gamma.

Moreover, the coarse Baum–Connes conjecture holds for the underlying metric space ∣Γ∣|\Gamma| associated to any proper left-invariant metric on Γ\Gamma, i.e. the coarse assembly map

μ∣Γ∣:lim⁡R→∞K∗(PR(∣Γ∣))⟶K∗(C∗(∣Γ∣))\mu_{|\Gamma|}:\lim_{R\to\infty}K_*(P_R(|\Gamma|))\longrightarrow K_*(C^*(|\Gamma|))

is an isomorphism.

We emphasize that our result establishes integral rather than merely rational injectivity of the Baum–Connes assembly map.

In Section 2, we introduce higher-order μ\mu-discreteness and higher-order orbit map. The proof of our main theorem fits into a more general construction that goes beyond Hilbert–Hadamard spaces. Thus, we shall first introduce a general type of spaces Ls0p(Ω,μ,E)L^p_{s_0}(\Omega,\mu,E), where EE is a snowflake metric bundle over Ω\Omega satisfying suitable assumptions, s0s_0 is a base section, and pp need not be 2. These spaces can be highly nonlinear and need not be Hilbert–Hadamard spaces. In § 3, we use an approach different from that of Gong, Wu, Xie and Yu [7] to obtain a more general result than Theorem 1.1. Our main theorem then follows immediately. In the final section, we present examples for which the first-order criterion fails but a second-order criterion applies.

AI disclosure. ChatGPT-6 was used solely for language editing and grammatical refinement during the preparation of this manuscript. All mathematical proofs and core results were developed by the authors.

Higher-order μ\mu-discreteness and higher-order orbit map

In this section, we first recall the notions of higher-order tangent bundles and higher-order derivatives. We then introduce LpL^p-direct integrals of metric bundles, which provide a common setting for higher-order μ\mu-discreteness and higher-order orbit map studied later. For a countable subgroup Γ<Diff⁡(M)\Gamma< \operatorname{Diff}(M), higher-order μ\mu-discreteness is defined by measuring how far the higher-order derivatives of elements of Γ\Gamma deviate from isometries.

Higher-order derivatives

The ordinary derivative map between tangent bundles captures only the first-order linearization of a smooth map: in local coordinates, it retains the linear term of the Taylor expansion and discards all quadratic and higher-order terms. In contrast, the rr-th order derivative map retains the full Taylor data through order rr. More precisely, its value at a point is determined by all partial derivatives of orders 1,…,r1,\ldots,r, and therefore encodes the entire Taylor expansion of a smooth map truncated at degree rr. Thus, at the first-order level one sees only infinitesimal linear behavior, whereas at higher-order levels one also sees the successive nonlinear terms of the Taylor expansion. This additional information allows us to study a broader class of subgroups of Diff⁡(M)\operatorname{Diff}(M). In this subsection, we review some standard constructions following [13 Section 12].

Definition 2.1. Let MM and NN be smooth manifolds, and let r≥0r \ge0 be an integer. Two smooth curves

ϕ,ψ:R⟶N\phi,\psi:\mathbb{R}\longrightarrow N

are said to have rr-th order contact at 00, written

ϕ∼rψ,\phi\sim_r\psi,

if, for every f∈C∞(N,R)f\in C^\infty(N,\mathbb{R}) and every k=0,1,…,rk=0,1,\ldots,r,

dkdtk∣t=0(f∘ϕ−f∘ψ)(t)=0.\left.\frac{d^k}{dt^k}\right|_{t=0}(f\circ\phi-f\circ\psi)(t)=0.

Obviously, ∼r\sim_r is an equivalence relation. Two smooth maps

ϕ,ψ:M⟶N\phi,\psi:M\longrightarrow N

are said to determine the same rr-jet at x∈Mx\in M if, for every smooth curve h:R→Mh:\mathbb{R}\to M with h(0)=xh(0)=x, the curves ϕ∘h\phi\circ h and ψ∘h\psi\circ h have rr-th order contact at zero, i.e.

ϕ∘h∼rψ∘h.\phi\circ h\sim_r\psi\circ h.

In this case, we write

jxrϕ=jxrψ.j_x^r\phi=j_x^r\psi.

The equivalence class jxrϕj_x^r\phi is called the rr-jet of ϕ\phi at xx, or an rr-jet of MM into NN. The set of all such rr-jets is denoted by

Jr(M,N).J^r(M,N).

The points xx and ϕ(x)\phi(x) are called the source and target of jxrϕ∈Jr(M,N)j_x^r\phi\in J^r(M,N) respectively. By Jxr(M,N)J_x^r(M,N) or Jr(M,N)yJ^r(M,N)_y, we mean the set of all rr-jets of MM into NN with source x∈Mx\in M or target y∈Ny\in N respectively, and we write

Jxr(M,N)y=Jxr(M,N)∩Jr(M,N)y.J_x^r(M,N)_y=J_x^r(M,N)\cap J^r(M,N)_y.

The space

TrM:=Jr(M,R)0T^rM:=J^r(M,\mathbb{R})_0

is called the space of all (1,r)(1,r)-covelocities on MM.

Since R\mathbb{R} is a vector space, Tr∗M⟶πMT^{r*}M \mathbin{\overset{\pi}{\longrightarrow}} M is a vector bundle, where π(jxrϕ)=x\pi(j_x^r\phi)=x and each fiber Txr∗M=Jxr(M,R)0T_x^{r*}M=J_x^r(M,\mathbb{R})_0 is equipped with the operations:

jxrϕ+jxrψ=jxr(ϕ+ψ)anda⋅jxrϕ=jxr(a⋅ϕ),a∈R.j_x^r\phi+j_x^r\psi=j_x^r(\phi+\psi)\quad\text{and}\quad a\cdot j_x^r\phi=j_x^r(a\cdot\phi),\quad a\in\mathbb{R}.

It is called the (1,r)(1,r)-covelocities bundle on MM.

Consider the dual vector bundle

T(r)M≔(Tr∗M)∗T^{(r)}M\coloneqq(T^{r*}M)^*

of the (1,r)(1,r)-covelocities bundle on MM. For every smooth map ϕ:M→N\phi:M\to N, the jet composition

A⟼A∘jxrϕ,x∈M,A∈Tϕ(x)r∗N,A\longmapsto A\circ j_x^r\phi,\quad x\in M,\quad A\in T_{\phi(x)}^{r*}N,

defines a linear map

λ(jxrϕ):Tϕ(x)r∗N⟶Txr∗M,\lambda(j_x^r\phi):T_{\phi(x)}^{r*}N\longrightarrow T_x^{r*}M,

and hence the dual map

(λ(jxrϕ))∗:Tx(r)M⟶Tϕ(x)(r)N.\bigl(\lambda(j_x^r\phi)\bigr)^*:T_x^{(r)}M\longrightarrow T_{\phi(x)}^{(r)}N.

We denote

Dx(r)ϕ=(λ(jxrϕ))∗.D_x^{(r)}\phi=\bigl(\lambda(j_x^r\phi)\bigr)^*.

They determine a smooth bundle map

D(r)ϕ:T(r)M⟶T(r)N.D^{(r)}\phi:T^{(r)}M\longrightarrow T^{(r)}N.

The assignments M↦T(r)MM\mapsto T^{(r)}M and ϕ↦D(r)ϕ\phi\mapsto D^{(r)}\phi define a covariant functor. In particular,

D(r)(ψ∘ϕ)=D(r)ψ∘D(r)ϕD^{(r)}(\psi\circ\phi)=D^{(r)}\psi\circ D^{(r)}\phi

References

  1. [1]F. Albiac and N. Kalton. Topics in Banach Space Theory, 2nd ed., Graduate Texts in Mathematics, Vol. 233, Springer, Cham, 2016.
  2. [2]P. Assouad. Plongements lipschitziens dans ℝⁿ, Bull. Soc. Math. France 111 (1983), 429–448.
  3. [4]M. Bridson and A. Haefliger. Metric spaces of non-positive curvature, Grundlehren der Math. Wiss., Vol. 319, Springer-Verlag, Berlin, 1999.DOI
  4. [5]A. Connes. Cyclic cohomology and the transverse fundamental class of a foliation. In Geometric Methods in Operator Algebras (Kyoto, 1983), Pitman Research Notes in Mathematics Series, Vol. 123, Longman Scientific & Technical, Harlow, 1986, pp. 52–144.
  5. [7]S. Gong, J. Wu, Z. Xie and G. Yu. The Novikov conjecture, the group of diffeomorphisms and continuous fields of Hilbert–Hadamard spaces, Mem. Eur. Math. Soc., to appear.
  6. [8]S. Gong, J. Wu and G. Yu. The Novikov conjecture, the group of volume preserving diffeomorphisms and Hilbert–Hadamard spaces, Geom. Funct. Anal. 31 (2021), 206–267.arxiv.org/abs/1811.02086
  7. [9]M. Gromov. Asymptotic invariants of infinite groups, in Geometric Group Theory, Vol. 2, G. A. Niblo and M. A. Roller (eds.), London Mathematical Society Lecture Note Series, Vol. 182, Cambridge Univ. Press, Cambridge, 1993, 1–295.
  8. [11]T. Hytönen, J. van Neerven, M. Veraar and L. Weis. Analysis in Banach Spaces. Volume I: Martingales and Littlewood–Paley Theory, Ergebnisse der Mathematik und ihrer Grenzgebiete, 3. Folge, Vol. 63, Springer, Cham, 2016.
  9. [12]G. Kasparov and G. Yu. The Novikov conjecture and geometry of Banach spaces, Geom. Topol. 16 (2012), 1859–1880.DOI
  10. [13]I. Kolář, P. W. Michor and J. Slovák. Natural Operations in Differential Geometry, Springer-Verlag, Berlin–Heidelberg, 1993.DOI
  11. [14]U. Lang and T. Schlichenmaier. Nagata dimension, quasisymmetric embeddings, and Lipschitz extensions, Int. Math. Res. Not. (2005), no. 58, 3625–3655.DOI
  12. [15]A. Naor and L. Silberman. Poincaré inequalities, embeddings, and wild groups, Compos. Math. 147 (2011), 1546–1572.
  13. [16]M. I. Ostrovskii. Embeddability of locally finite metric spaces into Banach spaces is finitely determined, Proc. Amer. Math. Soc. 140 (2012), 2721–2730.arxiv.org/abs/1103.0748
  14. [18]I. J. Schoenberg. Metric spaces and positive definite functions, Trans. Amer. Math. Soc. 44 (1938), 522–536.
  15. [19]G. Tian and G. Yu. Embedding complexity into the universal Banach space and the strong Novikov conjecture, arXiv:2605.12930.arxiv.org/abs/2605.12930
  16. [20]J. Wang, Z. Xie, G. Yu and B. Zhu. ℓ^p-coarse Baum–Connes conjecture for ℓ^q-coarse embeddable spaces, arXiv:2411.15070.arxiv.org/abs/2411.15070
  17. [24]G. Yu. The coarse Baum–Connes conjecture for spaces which admit a uniform embedding into Hilbert space, Invent. Math. 139 (2000), 201–240.DOI

Paper details

vlnc:W_djmxedwh
Version
v2 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents