nLab
homotopy dimension

Context

(,1)-Topos Theory

(∞,1)-topos theory

Background

Definitions

Characterization

Morphisms

Extra stuff, structure and property

Models

Constructions

structures in a cohesive (∞,1)-topos

Contents

Definition

Definition

An (∞,1)-topos 𝒳 has homotopy dimension n if every (n-1)-connected object A has a global element, a morphism *A from the terminal object into it.

This appears as HTT, def. 7.2.1.1.

Definition

An (∞,1)-topos 𝒳 is locally of homotopy dimension n if there exists a collection {U i𝒳} of objects such that

This appears as HTT, def. 7.2.1.8.

Properties

Proposition

If an (∞,1)-topos 𝒳 is locally of homotopy dimension n for some n then it is a hypercomplete (∞,1)-topos.

This appears as HTT, cor. 7.2.1.12.

Proposition

If 𝒳 has homotopy dimension n then it also has cohomological dimension n.

The converse holds if 𝒳 has finite homotopy dimension and n2.

This appears as HTT, cor. 7.2.2.30.

Proposition

An (∞,1)-topos 𝒳 has homotopy dimension n precisely if the global section (∞,1)-geometric morphism

(ΔΓ):𝒳ΓΔGrpd(\Delta \dashv \Gamma) : \mathcal{X} \stackrel{\overset{\Delta}{\leftarrow}}{\underset{\Gamma}{\to}} \infty Grpd

has the property that Γ sends (kn)-connective morphisms to (kn)-connective morphisms.

This is HTT, lemma 7.2.1.7

Examples

Proposition

Up to equivalence, the unique (∞,1)-topos of homotopy dimension 1 is the the terminal category *Sh (,1)().

This is HTT, example. 7.2.1.2.

Proof

An object X𝒳 is (1)-connected if the morphism X*to the terminal object in an (∞,1)-category is. This is the case if it is an effective epimorphism.

Since the global section (∞,1)-functor is corepresented by the terminal object, X is 0-connective precisely if Γ(X)Γ(*)=* is an epimorphism on connected components. By the discussion at effective epimorphism, this is the case precisely if Γ(X)* is an effective epimorphism in ∞Grpd.

So 𝒳 has homotopy dimension 0 if Γ preserves effective epimorphisms. This is the case if it preserves finit (∞,1)-limits (the (∞,1)-pullbacks defining a Cech nerve) and all (∞,1)-colimits (over the resulting Cech nerve). being a right adjoint (∞,1)-functor Γ always preserves (∞,1)-limits. If 𝒳 is local then Γ is by definition also a left adjoint and hence also preserves (∞,1)-colimits.

Proposition

Every local (∞,1)-topos has homotopy dimension 0.

Proof

Let

(ΔΓ):HGrpd(\Delta \dashv \Gamma \dashv \nabla) : \mathbf{H} \to \infty Grpd

be the terminal geometric morphism of the local (,1)-topos, with being the extra right adjoint to the global section (∞,1)-geometric morphism functor that characterizes locality.

By prop 3 it is sufficient to show that Γ send (-1)-connected morphisms to (-1)-connected morphisms, hence effective epimorphisms to effective epimorphisms.

By the existence of we have that Γ preserves not only (∞,1)-limits but also (∞,1)-colimits. Since effective epimorphisms are defined as certain colimits over diagrams of certain limits, Γ preserves effective epimorphisms.

So in particular for C any (∞,1)-category with a terminal object, the (∞,1)-category of (∞,1)-presheaves PSh (,1)(C) is an (∞,1)-topos of homotopy dimension 0. Notably Top ∞Grpd PSh (,1)(*) has homotopy dimension 0.

This is HTT, example. 7.2.1.3.

Proposition

Every (∞,1)-category of (∞,1)-presheaves is an (∞,1)-topos of local homotopy dimension 0.

This appears as HTT, example. 7.2.1.9.

Theorem

If a paracompact topological space X has covering dimension n, then the (∞,1)-category of (∞,1)-sheaves Sh (,1)(X):=Sh (,11)(Op(X)) is an (∞,1)-topos of homotopy dimension n.

This is HTT, theorem 7.2.3.6.

Proposition

For X ∞Grpd Top an object, the over-(∞,1)-topos Grpd/X has homotopy dimension n precisely if XTop a retract in the homotopy category Ho(Top) of a CW-complex of dimension n.

This is HTT, example 7.2.1.4.

References

The (∞,1)-topos theoretic notion is discuss in section 7.2.1 of

Revised on April 14, 2012 13:14:29 by Urs Schreiber (89.204.130.9)