nLab category of fibrant objects



Model category theory

model category, model \infty -category



Universal constructions


Producing new model structures

Presentation of (,1)(\infty,1)-categories

Model structures

for \infty-groupoids

for ∞-groupoids

for equivariant \infty-groupoids

for rational \infty-groupoids

for rational equivariant \infty-groupoids

for nn-groupoids

for \infty-groups

for \infty-algebras

general \infty-algebras

specific \infty-algebras

for stable/spectrum objects

for (,1)(\infty,1)-categories

for stable (,1)(\infty,1)-categories

for (,1)(\infty,1)-operads

for (n,r)(n,r)-categories

for (,1)(\infty,1)-sheaves / \infty-stacks

Homotopy theory

homotopy theory, (∞,1)-category theory, homotopy type theory

flavors: stable, equivariant, rational, p-adic, proper, geometric, cohesive, directed

models: topological, simplicial, localic, …

see also algebraic topology



Paths and cylinders

Homotopy groups

Basic facts




A category of fibrant objects is a category with weak equivalences equipped with extra structure somewhat weaker than that of a model category.

The extra structure of fibrations and cofibrations in a model category is, while convenient if it exists, not carried by many categories with weak equivalences which nevertheless admit many constructions in homotopy theory. Even if they do, sometimes the cofibrations are intractable in practice.

A category of fibrant objects is essentially like a model category but with all axioms concerning the cofibrations dropped, the concept of fibrations retained (“fibration category”) and assuming that all objects are fibrant (whence the name). It turns out that this is sufficient for many useful constructions. In particular, it is sufficient for giving a convenient construction of the homotopy category in terms of spans of length one. This makes categories of fibrant objects useful in homotopical cohomology theory.



A category of fibrant objects C\mathbf{C} is

  • a category with weak equivalences, i.e equipped with a subcategory

    Core(C)WC Core(\mathbf{C}) \hookrightarrow W \hookrightarrow C

    where fMor(W)f \in Mor(W) is called a weak equivalence;

  • equipped with a further subcategory

    Core(C)FC, Core(\mathbf{C}) \hookrightarrow F \hookrightarrow C \,,

    where fMor(F)f \in Mor(F) is called a fibration

Those morphisms which are both weak equivalences and fibrations are called acyclic fibrations.

This data has to satisfy the following properties:

  • CC has finite products, and in particular a terminal object *{*};

  • the pullback of a fibration along an arbitrary morphism exists, and is again a fibration;

  • acyclic fibrations are preserved under pullback;

  • weak equivalences satisfy 2-out-of-3

  • for every object there exists a path object

    • this means: for every object BB there exists at least one object denoted B IB^I (not necessarily but possibly the internal hom with an interval object) that fits into a diagram
      (BId×IdB×B)=(BσB Id 0×d 1B×B) (B \stackrel{Id \times Id}{\to} B \times B) = (B \stackrel{\sigma}{\to} B^I \stackrel{d_0 \times d_1}{\to} B \times B)

      where σ\sigma is a weak equivalence and d 0×d 1d_0 \times d_1 is a fibration;

  • all objects are fibrant, i.e. all morphisms B*B \to {*} to the terminal object are fibrations.


Full subcategories of model categories

The tautological example is the full subcategory of any model category on all objects which are fibrant.

Right proper model categories

Let MM be a right proper model category, let WW be the class of weak equivalences, and let F +F_+ be the class of morphisms ff in MM such that any pullback of ff in MM is also a homotopy pullback. Then MM together with WW and F +F_+ satisfy all the conditions to be a category of fibrant objects except possibly the condition that every morphism X*X \to {*} in MM is in F +F_+; so if we restrict to the full subcategory of those objects XX in MM such that X*X \to {*} is in F +F_+, then we do get a category of fibrant objects.

For example, sSet via its standard model structure is a category of fibrant objects in this way. The fibrations in this case are not the Kan fibrations (these also yields a category-of-fibrant-objects structure, via the above, but a different one) but are the sharp maps.


This includes notably all models for categories of infinity-groupoids:


The path object of any XX can be chosen to be the internal hom

X I=[Δ 1,X] X^I = [\Delta^1, X]

in with respect to the closed monoidal structure on SSet with the simplicial 1-simplex Δ 1\Delta^1.

The morphism XX IX \to X^I is given by the degeneracy map σ 0:Δ 0Δ 1\sigma_0 : \Delta^0 \to \Delta^1 as

X[Δ 0,X][σ 0,X][Δ 1,X]. X \stackrel{\simeq}{\to} [\Delta^0, X] \stackrel{[\sigma_0, X]}{\to} [\Delta^1, X] \,.

This is indeed a weak equivalence, since by the simplicial identities it is a section (a right inverse) for the morphism

[Δ 1,X][δ 0,X][Δ 0,X]. [\Delta^1, X] \stackrel{[\delta_0,X]}{\to} [\Delta^0, X] \,.

This map, one checks, has the right lifting property with respect to all boundary of a simplex-inclusions Δ nΔ n\partial \Delta^n \to \Delta^n. By a lemma discussed at Kan fibration this means that [δ 0,X][\delta_0,X] is an acyclic fibration. Hence [σ 0,X][\sigma_0, X], being its right inverse, is a weak equivalence.

The remaining morphism of the path space object X IX×XX^I \to X \times X is

[Δ 1,X][δ 0δ 1,X][Δ 0Δ 0,X]X×X. [\Delta^1, X] \stackrel{[\delta_0 \sqcup \delta_1, X]}{\to} [\Delta^0 \sqcup \Delta^0, X] \stackrel{\simeq}{\to} X \times X \,.

One checks that this is indeed a Kan fibration.

The stability of fibrations and acyclic fibrations follows from the above fact that both are characterized by a right lifting property (as described a model structure on simplicial sets).

See for instance section 1 of

  • Goerss, Jardine, Simplicial homotopy theory .

Concerning the example of Kan complexes, notice that SSet is also a category of co-fibrant objects (i.e. SSet opSSet^{op} is a category of fibrant objects) so that Kan complexes are in fact cofibrant and fibrant. That makes much of the technology discussed below superfluous, since it means that the right notion of \infty-morphism between Kan complexes is already the ordinary notion.

But then, often it is useful to model Kan complexes using the Dold-Kan correspondence, and then the second example becomes relevant, where no longer ever object is cofibrant.

Simplicial sheaves

The point of the axioms of a category of fibrant objects is that when passing from infinity-groupoids to infinity-stacks, i.e. to sheaves with values in infinity-groupoids, the obvious naïve way to lift the model structure from \infty-groupoids to sheaves of \infty-groupoids fails, as the required lifting axioms will be satisfied only locally (e.g. stalkwise).

One can get around this by employing a more sophisticated model category structure as described at model structure on simplicial presheaves, but often it is useful to use a more lightweight solution and consider sheaves with values in \infty-groupoids just as a category of fibrant objects, thereby effectively dispensing with the troublesome lifting property (as all mention of cofibrations is dropped):


(\infty-groupoid valued sheaves)

For CC be a site such that the sheaf topos Sh(C)Sh(C) has enough points, i.e. so that a morphism f:ABf : A \to B in Sh(X)Sh(X) is an isomorphism precisely if its image

x *f:x *Ax *B x^* f : x^* A \to x^* B

is a bijection of sets for all points (geometric morphisms from Sh(*)SetSh({*}) \simeq Set)

x:SetSh(*)x *x *Sh(C). x : Set \simeq Sh({*}) \stackrel{\stackrel{x^*}{\leftarrow}} {\stackrel{x_*}{\to}} Sh(C) \,.

Then let

C=SSh(C) \mathbf{C} = SSh(C)

be the full subcategory of

on those sheaves AA for which each stalk x *ASSetx^* A \in SSet is a Kan complex.

Define a morphism f:ABf : A \to B to be a fibration or a weak equivalence, if on each stalk x *f:x *Ax *Bx^* f : x^* A \to x^* B is a fibration or weak equivalence, respectively, of Kan complexes (in terms of the standard model structure on simplicial sets).


category of open subsets of XX.

The points of this topos precisely correspond to the ordinary points of XX.

Equipped with its structure as a category of fibrant objects, simplicial sheaves on XX are a model for infinity-stacks living over XX (the way an object ASh(X)A \in Sh(X) is a sheaf “over XX”).

  • Or let C=C = Diff be a (small model of) the site of smooth manifolds. The corresponding sheaf topos, that of smooth spaces has, up to isomorphism, one point per natural number, corresponding to the nn-dimensional ball D nD^n.

    Equipped with its structure as a category of fibrant objects, simplicial sheaves on DiffDiff are a model for smooth infinity-stacks.


SSh(X)SSh(X) with this structure is a category of fibrant objects.


The terminal object *=X{*} = X is the sheaf constant on the 0-simplex Δ 0\Delta^0, which represents the space XX itself as a sheaf.

For every simplicial sheaf AA and every point xXx \in X the stalk of the unique morphism A*A \to {*} is x *Ax *Xx^* A \to x^* {X}, which is the unique morphism from the Kan complex x *Ax^* A to Δ 0\Delta^0. Since Kan complexes are fibrant, this is a Kan fibration for every xXx \in X. So every AA is a fibrant object by the above definition.

The fact that fibrations and acyclic fibrations are preserved under pullback follows from the fact that the stalk operation

x *:SSh(X)SSh(pt)SSet x^* : SSh(X) \to SSh(pt) \simeq SSet

is the inverse image of a geometric morphism and hence preserves finite limits and in particular pullbacks. So if f:ABf : A \to B is a fibration or acyclic fibration in SSh(X)SSh(X) and

A× BC B h *f f C h B \array{ A \times_B C &\to& B \\ \downarrow^{\mathrlap{h^* f}} && \downarrow^\mathrlap{f} \\ C &\stackrel{h}{\to}& B }

is a pullback diagram in SSh(X)SSh(X), then for xXx \in X any point of XX also

x *(A× BC) x *B x *(h *f) x *f x *C x * B \array{ x^*(A \times_{B} C) &\to& x^*B \\ \downarrow^{\mathrlap{x^* (h^* f)}} && \downarrow^{\mathrlap{x^* f}} \\ x^*C &\stackrel{x^*}{\to}& B }

is a pullback diagram, now of Kan complexes. Since Kan complexes form a category of fibrant objects, by the above, it follows that x *(h *f)x^* (h^* f) is a fibration or acyclic fibration of Kan complexes, respectively. Since this holds for every xx, it follows that h *fh^* f is a fibration or acyclic fibration, respectively, in SSh(X)SSh(X).

Recall that a functorial choice of path object for a Kan complexe KK is the internal hom [Δ 1,K][\Delta^1, K] with respect to the closed monoidal structure on simplicial sets:

K=[Δ 0,K]s 0[Δ 1,K]d 0×d 1[Δ 0,K]×[Δ 0,K]=K×K, K \stackrel{=}{\to} [\Delta^0, K] \stackrel{s_0}{\to} [\Delta^1, K] \stackrel{d_0 \times d_1}{\to} [\Delta^0, K] \times [\Delta^0, K] \stackrel{=}{\to} K \times K \,,

where s is_i and d id_i denote the degeneracy and face maps, respectively.

For ASSh(X)A \in SSh(X) let [Δ 1,A][\Delta^1,A] denote the sheaf

[Δ 1,A]:U[Δ 1,A(U)], [\Delta^1,A] : U \mapsto [\Delta^1,A(U)] \,,

where on the left we have new notation and on the right we have the internal hom in SSet.

(The notation on the left defines the way in which SSh(X)SSh(X) is copowerered over SSet).

We want to claim that [Δ 1,A][\Delta^1,A] is a path object for AA.

To check that [Δ 1,A][\Delta^1,A] is fibrant, let xXx \in X be any point and consider the stalk x *[Δ 1,A]SSetx^* [\Delta^1,A] \in SSet. We compute laboriously

x *[Δ 1,A] colim Ux[Δ 1,A(U)] colim UxSSet(Δ 1×Δ ,A(U)) ([n]colim UxSSet(Δ 1×Δ n,A(U)) ([n]colim Ux [k]ΔSet(Δ([k],[1])×Δ([k],[n]),A(U) k) ([n] [kn+1]Δ(colim UxSet(Δ([k],[1])×Δ([k],[n]),A(U) k) ([n] [k]Δ| n+1(Set(Δ([k],[1])×Δ([k],[n]),colim UxA(U) k) ([n] [k]Δ| n+1(Set(Δ([k],[1])×Δ([k],[n]),(colim UxA(U)) k) [Δ 1,colim UxA(U)] [Δ 1,x *A]) \begin{aligned} x^* [\Delta^1,A] &\simeq colim_{U \ni x} [\Delta^1,A(U)] \\ &\simeq colim_{U \ni x} SSet(\Delta^1 \times \Delta^\bullet, A(U)) \\ &\simeq ([n] \mapsto colim_{U \ni x} SSet(\Delta^1 \times \Delta^\n, A(U)) \\ & \simeq ([n] \mapsto colim_{U \ni x} \int_{[k] \in \Delta} Set(\Delta([k],[1])\times\Delta([k],[n]), A(U)_k) \\ & \simeq ([n] \mapsto \int_{[k \leq n+1] \in \Delta}( colim_{U \ni x} Set(\Delta([k],[1])\times\Delta([k],[n]), A(U)_k ) \\ &\simeq ([n] \mapsto \int_{[k] \in \Delta|_{n+1}}( Set(\Delta([k],[1])\times\Delta([k],[n]), colim_{U \ni x} A(U)_k ) \\ &\simeq ([n] \mapsto \int_{[k] \in \Delta|_{n+1}}( Set(\Delta([k],[1])\times\Delta([k],[n]), (colim_{U \ni x} A(U))_k ) \\ &\simeq [\Delta^1, colim_{U \ni x} A(U)] \\ & \simeq [\Delta^1, x^* A] ) \end{aligned}

Where the

  • first step is the general formula for the stalk;

  • second step is the formula for the internal hom in the closed monoidal structure on simplicial sets;

  • third step is the fact that colimits of presheaves are computed objectwise (see examples at colimit);

  • the fourth step is the definition of the SSet-enriched functor category by an end

  • the fifth step uses that

  • the sixth step uses that the set Δ([k],[1])×Δ([k],[n])\Delta([k],[1])\times \Delta([k],[n]) is finite, hence a compact object so that the colimit can be taken into the hom;

  • the seventh step uses again that colimits of presheaves are computed objectwise

  • the remaining steps then just rewind the first ones, only that now A(U)A(U) has been replaced by colim UxA(U)colim_{U \ni x} A(U).

That the morphism A[Δ 1,A]A \to [\Delta^1,A] is a weak equivalence and that [Δ 1,A]d 0×d 1A×A[\Delta^1,A] \stackrel{d_0 \times d_1}{\to} A \times A is a fibration follows similarly by taking the stalk colimit inside to reduce to the statement that x *A[Δ 1,x *A]x^* A \to [\Delta^1,x^* A] is a weak equivalence and [Δ 1,x *A]d 0×d 1x *A×x *A[\Delta^1,x^* A] \stackrel{d_0 \times d_1}{\to} x^*A \times x^* A is a fibration, using that [Δ 1,x *A][\Delta^1,x^*A] is a path object for the Kan complex x *Ax^* A.

The category of fibrant objects SSh(X)SSh(X) is in fact the motivating example in BrownAHT. Notice that the homotopy category in question coincides with that using the model structure on simplicial presheaves, so that the category of fibrant objects of stalk-wise Kan sheaves is a model for the homotopy category of infinity-stacks.


Let GG be a topogical group and recall that BG\mathbf{B} G denotes the corresponding one-object groupoid.

For XX a topological space and UU an open subset, let C(U,G)SetC(U, G) \in Set be the set of continuous maps from UU into GG. This set naturally is itself a group, so that to each UXU \subset X we may associuate the one-object groupoid

UBC(U,G). U \mapsto \mathbf{B} C(U,G) \,.

By postcomposition this with the nerve operation we obtain an assignment of Kan complexes to open subsets:

UNBC(U,G). U \mapsto N \mathbf{B} C(U,G) \,.

In degree 0 this is the constant sheaf

(NB(,G)) 0:U* (N \mathbf{B}(-,G))_0 : U \mapsto {*}

while in degree 1 this is the sheaf of GG-valued functions

(NB(,G)) 1:UC(U,G). (N \mathbf{B}(-,G))_1 : U \mapsto C(U,G) \,.

When the context is understood, we will just write BG\mathbf{B}G again for this \infty-groupoid valued sheaf

BG:=NBC(,G). \mathbf{B}G := N \mathbf{B} C(-,G) \,.

Slice categories

Let C\mathbf{C} be a category of fibrant objects, with fibrations FMor(C)F \subset Mor(\mathbf{C}) and weak equivalences WMor(C)W \subset Mor(\mathbf{C}).

For any object BB in C\mathbf{C}, let C B F\mathbf{C}_B^F be the category of fibrations over BB (a full subcategory of the slice category C/B\mathbf{C}/B):

  • objects are fibrations ABA \to B in C\mathbf{C},

  • morphisms are commuting triangles

    A A F F B \array{ A &&\to&& A' \\ & {}_{\in F}\searrow && \swarrow_{\in F} \\ && B }

    in C\mathbf{C}.

There is an obvious forgetful functor C B FC\mathbf{C}_B^F \to \mathbf{C}, which induces notions of weak equivalence and fibration in C B F\mathbf{C}_B^F.


With this structure, C B F\mathbf{C}_B^F becomes a category of fibrant objects.


Below is proven the factorization lemma that holds in any category of fibrant objects. This implies in particular that every morphism

A Id×Id A× BA F F B \array{ A &&\stackrel{Id \times Id}{\to}&& A \times_B A \\ & {}_{\in F}\searrow && \swarrow_{\in F} \\ && B }

may be factored as

A W A I F A× BA F F B. \array{ A &\stackrel{\in W}{\to}& A^I &\stackrel{\in F}{\to}& A \times_B A \\ & {}_{\in F}\searrow && \swarrow_{\in F} \\ && B } \,.

This provides the path space objects in C B F\mathbf{C}^F_B.


Simple consequences of the definition

Before looking at more sophisticated constructions, we record the following direct consequences of the definition of a category of fibrant objects.


For every two objects A 1,A 2CA_1, A_2 \in \mathbf{C}, the two projection maps

p i:A 1×A 2FA i p_i : A_1 \times A_2 \stackrel{\in F}{\to} A_i

out of their product are fibrations.


Because by assumption both morphisms A i*A_i \to {*} are fibrations and fibrations are preserved under pullback

A 1×A 2 A 2 p 1 F F A 1 *. \array{ A_1 \times A_2 &\to& A_2 \\ \;{}^{p_1}\downarrow^{\mathrlap{\Rightarrow \in F}} && \downarrow^{\mathrlap{\in F}} \\ A_1 &\to& {*} } \,.

For every object BCB \in \mathbf{C} and everey path object B IB^I of BB, the two morphisms

d i:B IWFB d_i : B^I \stackrel{\in W \cap F}{\to} B

(whose product d 0×d 1d_0 \times d_1, recall, is required to be a fibration) are each separately acyclic fibrations.


By the above lemma d i:B Id 0×d 1B×Bp iBd_i : B^I \stackrel{d_0 \times d_1}{\to} B \times B \stackrel{p_i}{\to} B is the composite of two fibrations and hence itself a fibration.

Moreover, from the diagram

B B I d 0×d 1 B×B d i p i Id B \array{ B &\stackrel{\simeq}{\to}& B^I &\stackrel{d_0 \times d_1}{\to}& B \times B \\ &&&\searrow^{d_i}& \downarrow^{\mathrlap{p_i}} \\ & \searrow^{Id}&&& B }

one reads off that the 2-out-of-3 property for weak equivalences implies that d id_i is also a weak equivalence.

Generalized universal bundles and the factorization lemma

A central lemma in the theory of categories of fibrant objects is the following factorization lemma.

Lemma (factorization lemma)

For every morphism f:CDf : C \to D in a category C\mathbf{C} of fibrant objects, there is an object E fB\mathbf{E}_f B such that ff factors as

E fB σ fW p fF C f B \array{ && \mathbf{E}_f B \\ & {}^{\sigma_f \in W}\nearrow && \searrow^{p_f \in F} \\ C &&\stackrel{f}{\to}&& B }


  • p fp_f a fibration

  • σ f\sigma_f a weak equivalence that is a section ( a right inverse) of an acyclic fibration:

    Id E fB=(Cσ fE fBC). Id_{\mathbf{E}_f B} = ( C \stackrel{\sigma_f}{\to} \mathbf{E}_f B \stackrel{\simeq}{\to} C ) \,.

This is the analog of one of the factorization axioms in a model category which says that every map factors as an acyclic cofibration followed by a fibration.

Notice that by 2-out-of-3 this in particular implies that every weak equivalence f:CWBf : C \stackrel{\in W}{\to} B is given by a span of acyclic fibrations.

E fB σ fW p fFW C fW B. \array{ && \mathbf{E}_f B \\ & {}^{\sigma_f \in W}\nearrow && \searrow^{p_f \in F\cap W} \\ C &&\stackrel{f \in W}{\to}&& B } \,.

In the context of Lie groupoid theory these are known as the Morita equivalences between groupoids. There here arise as a special case. Compare also the notion of anafunctor.

The way the proof of this lemma works, one sees that this really arises in the wider context of computing homotopy pullbacks in CC. Therefore we split the proof in two steps that are useful in their own right and will be taken up in the next section on homotopy limits.


For f:CBf : C \to B a morphism in C\mathbf{C}, we say that the morphism p f:E fBBp_f : \mathbf{E}_f B \to B defined as the composite vertical morphism in the pullback diagram

E fB C f B I d 0 B d 1 B \array{ \mathbf{E}_f B &\stackrel{\simeq}{\to}& C \\ \downarrow && \downarrow^{\mathrlap{f}} \\ B^I &\stackrel{d_0}{\to}& B \\ \downarrow^{\mathrlap{d_1}} \\ B }

for some path space object B IB^I is the generalized universal bundle over BB relative to ff.

The universal bundle terminology is best understood from the following example


Consider the category of fibrant objects given by Kan complexes or just strict omega-groupoids.

For GG an ordinary group write BG\mathbf{B} G for the corresponding groupoid. When regarding GG as a constant simplicial group the corresponding Kan complex is often denoted W¯G\bar W G (see simplicial group) but we shall just write BG\mathbf{B} G also for this Kan complex, for simplicity.

The corresponding path object is given by the groupoid (or its corresponding Kan complex)

(BG) I=[Δ 1,BG]=G\\G//G (\mathbf{B} G)^I = [\Delta^1, \mathbf{B} G ] = G \backslash \backslash G//G

where the right denotes the action groupoid of G×GG \times G acting on GG by left and right multiplication.

Let *:*BG{*} : {*} \to \mathbf{B} G be the unique morphism from the point into BG\mathbf{B} G. The corresponding generalized universal bundle is

E *G=G//G \mathbf{E}_{*} G = G//G

the action groupoid of GG acting on itself from just the right. (The corresponding Kan complex is traditionally denoted WGW G when thought of as a simplicial group).

That G//GBGG//G \to \mathbf{B}G is indeed the universal GG-principal bundle (under the Quillen equivalence of Kan complexes and topological spaces) is an old result of Segal (as described at generalized universal bundle).


The morphism p f:E fBBp_f : \mathbf{E}_f B \to B is a fibration.


The defining pullback diagram for E fB\mathbf{E}_f B can be refined to a double pullback diagram as follows

E fB F C×B p 1 C f×Id f B I d 0×d 1F B×B p 1 B d 1 p 2F B. \array{ \mathbf{E}_f B &\stackrel{\in F}{\to}& C \times B &\stackrel{p_1}{\to}& C \\ \downarrow && \downarrow^{\mathrlap{f \times Id}} && \downarrow^{\mathrlap{f}} \\ B^I &\stackrel{d_0 \times d_1 \in F}{\to}& B \times B &\stackrel{p_1}{\to}& B \\ \downarrow^{\mathrlap{d_1}} & \swarrow_{p_2 \in F} \\ B } \,.

Both squares are pullback squares. Since pullbacks of fibrations are fibrations, the morphism E fBC×B\mathbf{E}_f B \to C \times B is a fibration.

By one of the lemmas above, also the projection map p i:B×BBp_i : B \times B \to B is a fibration.

The above diagram exibits p fp_f as the the composite

p f :E fBC×Bf×IdB×Bp 2B =E fBC×Bp 2B \begin{aligned} p_f &: \mathbf{E}_f B \to C \times B \stackrel{f \times Id}{\to} B \times B \stackrel{p_2}{\to} B \\ & = \mathbf{E}_f B \to C \times B \stackrel{p_2}{\to} B \end{aligned}

of two fibrations. Therefore it is itself a fibration.


The morphism E fBC\mathbf{E}_f B \stackrel{\simeq}{\to} C has a section (a right inverse) σ f:CE fB\sigma_f : C \stackrel{\simeq}{\to} \mathbf{E}_f B and its composite with p fp_f is ff:

E fB σ f C p f f B \array{ \mathbf{E}_f B &\stackrel{\sigma_f}{\leftarrow}&& C \\ \downarrow^{\mathrlap{p_f}} && \swarrow_{f} \\ B }

The section

σ f=Id×σf \sigma_f = Id \times \sigma \circ f

is the morphism induced via the universal property of the pullback by the section σ:BB I\sigma : B \to B^I of d 0:B IBd_0 : B^I \to B:

C σ fW E fB WF C f f B σ B I d 1WF B Id d 0 B=C Id C f f B Id B. \array{ C &\stackrel{\sigma_f \in W}{\to}& \mathbf{E}_f B &\stackrel{\in W \cap F}{\to}& C \\ \downarrow^{\mathrlap{f}} && \downarrow && \downarrow^{\mathrlap{f}} \\ B &\stackrel{\sigma}{\to}& B^I &\stackrel{d_1 \in W \cap F}{\to}& B \\ & {}_{Id}\searrow & \downarrow^{\mathrlap{d_0}} \\ && B } \;\;\;\; = \;\;\;\; \array{ C &\stackrel{Id}{\to}& C \\ \downarrow^{\mathrlap{f}} && \downarrow^{\mathrlap{f}} \\ B &\stackrel{Id}{\to}& B } \,.

More sophisticated consequences of the definition

Using the factorization lemma, one obtaines the following further useful statements about categories of fibrant objects:

Recall that plain weak equivalences, if they are not at the same time fibrations, are not required by the axioms to be preserved by pullback. But it follows from the axioms that weak equivalences are preserved under pullback along fibrations.

This we establish in two lemmas.



A 1 f A 2 F F B \array{ A_1 &&\stackrel{f}{\to}&& A_2 \\ & {}_{\in F}\searrow && \swarrow_{\in F} \\ && B }

be a morphism of fibrations over some object BB in C\mathbf{C} and let u:BBu : B' \to B be any morphism in C\mathbf{C}. Let

u *A 1 u *f u *A 2 F F B \array{ u^*A_1 &&\stackrel{u^* f}{\to}&& u^* A_2 \\ & {}_{\in F}\searrow && \swarrow_{\in F} \\ && B' }

be the corresponding morphism pulled back along uu.


  • if fFf \in F then also u *fFu^* f \in F;

  • if fWf \in W then also u *fWu^* f \in W.


For fFf \in F the statement follows from the fact that in the diagram

B× BA 1 A 1 u *fF fF B× BA 2 A 2 F F B u B \array{ B' \times_B A_1 &\to& A_1 \\ \;\;\downarrow^{\mathrlap{u^* f \in F}} && \;\;\downarrow^{\mathrlap{f \in F}} \\ B' \times_B A_2 &\to& A_2 \\ \;\downarrow^{\mathrlap{\in F}} && \;\downarrow^{\mathrlap{\in F}} \\ B' &\stackrel{u}{\to}& B }

all squares (the two inner ones as well as the outer one) are pullback squares, since pullback squares compose under pasting.

The same reasoning applies for fWFf \in W \cap F.

To apply this reasoning to the case where fWf \in W, we first make use of the factorization lemma to decompose ff as a right inverse to an acyclic fibration followed by an acyclic fibration.

f:A 1WE fA 2WFA 2. f : A_1 \stackrel{\in W}{\to} \mathbf{E}_f A_2 \stackrel{\in W \cap F}{\to} A_2 \,.

(Compare the definition of the category of fibrant objects C B F\mathbf{C}_B^F of fibrations over BB, discussed in the example section above.)

Using the above this reduces the proof to showing that the pullback of the top horizontal morphism of

A 1 E fA 2 F F B \array{ A_1 &&\stackrel{}{\to}&& \mathbf{E}_f A_2 \\ & {}_{\in F}\searrow && \swarrow_{\in F} \\ && B }

(here the fibration on the right is the composite of the fibration E fA 2A 2\mathbf{E}_f A_2 \to A_2 with A 2BA_2 \to B)

along uu is a weak equivalence. For that consider the diagram

B× BA 1 A 1 B× BE fA 2 E fA 2 WF WF B× BA 1 A 1 F F B B \array{ B' \times_B A_1 &\to& A_1 \\ \downarrow && \downarrow \\ B' \times_B \mathbf{E}_f A_2 &\to& \mathbf{E}_f A_2 \\ \;\;\downarrow^{\mathrlap{\in W \cap F}} && \;\;\downarrow^{\mathrlap{\in W \cap F}} \\ B' \times_B A_1 &\to& A_1 \\ \;\downarrow^{\mathrlap{\in F}} && \;\downarrow^{\mathrlap{\in F}} \\ B' &\to& B }

where again all squares are pullback squares. The top two vertical composite morphisms are identities. Hence by 2-out-of-3 the morphism B× BE 1B× BEB' \times_B E_1 \to B' \times_B \mathbf{E} is a weak equivalence.


The pullback of a weak equivalence along a fibration is again a weak equivalence.


Let be a fibration p:EBp:E\to B and let e:ABe:A\to B be a weak equivalence. Without loss of generality we can assume that ee is a section of an acyclic fibration q:BAq:B\to A – otherwise we first decompose with the factorization lemma and then pull back along the factors individually.

Consider the following diagram where the four tiled squares are pullbacks, f=p,idf=\langle{p,\mathrm{id}}\rangle, and the bent back rectangle is also a pullback. ll is a weak equivalence since qq is a acyclic fibration, and by 2-out-of-3 we can conclude that ff is a weak equivalence as well. By the preceding lemma , so is gg. The goal is to show that hh is a weak equivalence, and by 2-out-of-3 it’s enough to check that for kk. This follows again from 2-out-of-3 since kk is a section of ll (because the bottom row of the diagram is an identity).


Model categories

that satisfy this property are called right proper model categories.

Right properness is a crucial assumption in the closely related work

  • Jardine, Cocycle categories (arXiv)

Homotopy fiber product

Using the existence of path space objects one can construct specific homotopy pullbacks called homotopy fiber products .


A homotopy fiber product or homotopy pullback of two morphisms

AuCvB A \stackrel{u}{\to} C \stackrel{v}{\leftarrow} B

in a category of fibrant objects is the object A× CC I× CBA \times_C C^I \times_C B defined as the (ordinary) limit

A× CC I× CB B v C I d 0 C d 1 A u C. \array{ A \times_C C^I \times_C B &&&\to & B \\ &&&& \downarrow^v \\ & &C^I & \stackrel{d_0}{\to}& C \\ \downarrow && \downarrow^{\mathrlap{d_1}} \\ A &\stackrel{u}{\to} & C } \,.

This essentially says that A× CC I× CBA \times_C C^I \times_C B is the universal object that makes the diagram

A× CC I× CB B v A u C \array{ A \times_C C^I \times_C B &\to& B \\ \downarrow && \downarrow^{\mathrlap{v}} \\ A &\stackrel{u}{\to}& C }

commute up to homotopy (see the section on homotopies for more on that).


These homotopy pullbacks present indeed the correct (infinity,1)-limits, this is the content of prop. below.


The projection

A× CC I× CBA A \times_C C^I \times_C B \to A

out of a homotopy fiber product is a fibration. If v:BCv : B \to C is a weak equivalence, then this is an acyclic fibration.

The same is of course true for the map to BB and the morphism u:ACu : A \to C, by symmetry of the diagram.


One may compute this limit in terms of two consecutive pullbacks in two different ways.

On the one hand we have

A× CC I× CB E vC B v C I d 0 C d 1 A u C \array{ A \times_C C^I \times_C B &\to& \mathbf{E}_v C &\to & B \\ && \downarrow && \downarrow^{\mathrlap{v}} \\ & &C^I & \stackrel{d_0}{\to}& C \\ \downarrow && \downarrow^{\mathrlap{d_1}} \\ A &\stackrel{u}{\to}& C }

where both squares are pullback squares.

By the above lemma on generalized universal bundles, the map E vCC\mathbf{E}_v C \to C is a fibration. The first claim follows then since fibrations are stable under pullback.

On the other hand we can rewrite the limit diagram also as

A× CC I× CB B v E uC WF C I d 0WF C WF d 1WF A u C \array{ A \times_C C^I \times_C B &\to& && B \\ \downarrow && && \downarrow^{\mathrlap{v}} \\ \mathbf{E}_u C & \stackrel{\in W \cap F}{\to} &C^I & \stackrel{d_0 \in W \cap F}{\to}& C \\ \downarrow^{\mathrlap{\in W \cap F}} && \;\;\downarrow^{\mathrlap{d_1\in W \cap F}} \\ A &\stackrel{u}{\to} & C }

where again both inner squares are pullback squares.

Again by the above statement on generalized universal bundles, we have that the morphism E uCC\mathbf{E}_u C \to C is a fibration. By one of the above propositions, weak equivalences are stable under pullback along fibrations, hence the pullback A× CC I× CBE uCA \times_C C^I \times_C B \to \mathbf{E}_u C of vv is a weak equivalence. Since also E uCA\mathbf{E}_u C \to A is a weak equivalence (being the pullback of an acyclic fibration) the entire morphism A× CC I× CBAA \times_C C^I \times_C B \to A is.



Two morphism f,g:ABf,g : A \to B in C(A,B)C(A,B) are

  • right homotopic, denoted fgf \simeq g, precisely if they fit into a diagram

    B f d 0 A η B I g d 1 B \array{ && B \\ & {}^f\nearrow & \uparrow^{d_0} \\ A &\stackrel{\eta}{\to}& B^I \\ & {}_g\searrow & \downarrow^{\mathrlap{d_1}} \\ && B }

    for some path space object B IB^I;

  • homotopic, denoted fgf \sim g, if they become right homotopic after pulled back to a weakly equivalent domain, i.e. precisely if they fit into a diagram

A f B wW d 0 A^ η B I wW d 1 A g B \array{ && A &\stackrel{f}{\to}& B \\ &{}^{w \in W}\nearrow&&& \uparrow^{d_0} \\ \hat A && \stackrel{\eta}{\to} && B^I \\ &{}_{w\in W}\searrow & && \downarrow^{d_1} \\ && A &\stackrel{g}{\to}& B }

for some object A^\hat A and for some path space object B IB^I of II


So this says that there is a right homotopy between the two morphisms after both are pulled back to a sufficiently good resolution of their domain.


For A,BCA,B \in \mathbf{C}, right homotopy is an equivalence relation on the hom-set C(A,B)\mathbf{C}(A,B).


This follows by “piecing path spaces together”:

Let B I 1B^{I_1} and B I 2B^{I_2} be two path space objects of BB. Then the pullback

B I 1I 2 B I 2 d 0 B I 1 d 1 B \array{ B^{I_1 \vee I_2} &\to& B^{I_2} \\ \downarrow && \downarrow^{d_0} \\ B^{I_1} &\stackrel{d_1}{\to}& B }

defines a new path object, with structure maps

Bσ 1×σ 2B I 1I 2(d 0p 1)×(d 1p 2)B×B. B \stackrel{\sigma_1 \times \sigma_2}{\to} B^{I_1 \vee I_2} \stackrel{(d_0 \circ p_1) \times (d_1\circ p_2)}{\to} B \times B \,.

So given two right homotopies with respect to B I 1B^{I_1} and B i 2B^{i_2} we can paste them next to each other and deduce a homotopy through B I 1I 2B^{I_1 \vee I_2}

B f d 0 1 A η 1 B I 1 g d 1 1 B B I 1I 2 g d 0 2 A η 2 B I 2 h d 1 2 B \array{ && B \\ & {}^f\nearrow & \uparrow^{d_0^1} \\ A &\stackrel{\eta_1}{\to}& B^{I_1} \\ & {}_{g}\searrow& \downarrow^{\mathrlap{d_1^1}} & \nwarrow \\ && B && B^{I_1 \vee I_2} \\ & {}^{g}\nearrow & \downarrow^{\mathrlap{d_0^2}} & \swarrow \\ A &\stackrel{\eta_2}{\to}& B^{I_2} \\ & {}_h\searrow & \downarrow^{\mathrlap{d_1^2}} \\ &&B }

We next similarly want to deduce that not only right homotopy fgf \simeq g but also true homtopy fgf \sim g defines an equivalence relation on hom-sets C(A,B)\mathbf{C}(A,B). For that we need the following to lemmas.


Every diagram

A E iW pF X B \array{ A &\to& E \\ \;\;\downarrow^{\mathrlap{i \in W}} && \;\;\downarrow^{\mathrlap{p \in F}} \\ X &\to& B }

may be refined to a diagram

A X E i tWF pF X B \array{ A &\to & X' &\to& E \\ & {}_{i}\searrow & \;\;\downarrow^{\mathrlap{t \in W \cap F}} && \;\;\downarrow^{\mathrlap{p \in F}} \\ && X &\to& B }

Consider the pullback square

A X× BE E iW F F X B \array{ A &\to& X \times_B E &\to& E \\ &{}_{i \in W}\searrow& \;\; \downarrow^{\mathrlap{\in F}} && \;\; \downarrow^{\mathrlap{\in F}} \\ && X &\to& B }

and apply the factorization lemma, lemma , to factor the universal morphism AX× BEEA \to X \times_B E \to E into the pullback as

AWEEFE A \stackrel{\in W}{\to} \mathbf{E} E \stackrel{\in F}{\to} E

to obtain the diagram

A EE E iW F F X B, \array{ A &\stackrel{\simeq}{\to}& \mathbf{E} E &\to& E \\ &{}_{i \in W}\searrow& \;\; \downarrow^{\mathrlap{\in F}} && \;\; \downarrow^{\mathrlap{\in F}} \\ && X &\to& B } \,,

where the middle vertical morphism is still a fibration, being the composite of two fibrations. By 2-out-of-3 it follows that it is also a weak equivalence.


For u:BC u : B \to C a morphism and B IB^I, C IC^I choices of path objects, there is always another path object B IB^{I'} with an acyclic fibration B IWFB IB^I \stackrel{\in W \cap F}{\leftarrow} B^{I'} and a span of morphisms of path space objects

B = B u C σ σ σ C B I WF B I u¯ C I d 0×d 1 d 0×d 1 d 0 C×d 1 C B×B = B×B u×u C×C \array{ B &\stackrel{=}{\leftarrow}& B &\stackrel{u}{\to}& C \\ \downarrow^{\mathrlap{\sigma}} && \downarrow^{\sigma'} && \downarrow^{\mathrlap{\sigma_C}} \\ B^I &\stackrel{\in W \cap F}{\leftarrow}& B^{I'} &\stackrel{\bar u}{\to}& C^I \\ \;\;\downarrow^{d_0 \times d_1} && \;\;\downarrow^{\mathrlap{d'_0 \times d'_1}} && \;\;\downarrow^{\mathrlap{d_0^C \times d_1^C}} \\ B \times B &\stackrel{=}{\leftarrow}& B \times B &\stackrel{u \times u}{\to}& C \times C }

Apply the lemma above to the square

B u C σ C C I σ d 0×d 1 B I d 0×d 1 B×B u×u C×C. \array{ B &\stackrel{u}{\to}& C &\stackrel{\sigma_C}{\to}& C^I \\ \downarrow^{\mathrlap{\sigma}} &&&& \downarrow^{\mathrlap{d_0 \times d_1}} \\ B^I &\stackrel{d_0 \times d_1}{\to}& B \times B &\stackrel{u \times u}{\to}& C \times C } \,.

Right homotopy fgf \simeq g between morphisms is preserved under pre- and postcomposition with a given morphism.

More precisely, let f,g:BCf, g : B \to C be two homotopic morphisms. Then

  • for all morphisms ABA \to B and CDC \to D the composites ABfCDA \to B \stackrel{f}{\to} C \to D and ABgCDA \to B \stackrel{g}{\to} C \to D are still right homotopic.

  • moreover, the right homotopy may be realized with every given choice of
    path space object D ID^I for DD.


We decompose this into two statements:

  1. for any ABA \to B the morphisms ABf,gBA \to B \stackrel{f,g}{\to} B are right homotopic.

  2. for any u:CDu : C \to D and choice D ID^I of path object there is an acyclic fibration BB B' \to B such that BBfCDB' \to B \stackrel{f}{\to} C \to D is right homotopic to BBgCDB' \to B \stackrel{g}{\to} C \to D by a right homotopy η:BD I\eta : B' \to D^I.

The first of these follows trivially.

The second one follows by using the weak functoriality property of path objects from above: let B:=B× C IC IB' := B \times_{C^I} C^{I'} be the pullback in the following diagram

B C I u¯ D I WF WF B η C I f×g C×C u×u D×D \array{ B' &\to& C^{I'} &\stackrel{\bar u}{\to}& D^I \\ \;\;\;\downarrow^{\mathrlap{\in W \cap F}} && \;\;\;\downarrow^{\mathrlap{\in W \cap F}} && \downarrow \\ B &\stackrel{\eta}{\to}& C^I \\ &{}_{f \times g}\searrow & \downarrow && \downarrow \\ && C \times C &\stackrel{u \times u}{\to}& D \times D }

We need one more intermediate result for seeing that homotopy is an equivalence relation

  • Every diagram

    B wW A C \array{ && B \\ && \downarrow^{\mathrlap{w \in W}} \\ A &\to& C }

    in C\mathbf{C} extends to a (right) homtopy-commutative diagram

    A B wW wW A C. \array{ A' &\to & B \\ \downarrow^{\mathrlap{w' \in W}} && \downarrow^{\mathrlap{w \in W}} \\ A &\to& C } \,.
  • For every pair of morphisms

    f,g,AB f, g, A \stackrel{\to}{\to} B

    and weak equivalence t:BWCt : B \stackrel{\in W}{\to} C such that there is a right homotopy tftgt \circ f \simeq t \circ g, there exists a weak equivalence t:AAt' : A' \to A such that ftgtf \circ t' \simeq g \circ t'.

  • The first point we accomplish this by letting A:=A× CC I× CBA' := A \times_C C^I \times_C B be the homotopy fiber product in CC of a representative of the pullback diagram. The lemma about morphisms out of the homotopy fiber product says that AAA' \to A is a weak equivalence.

  • The second point is more work. Let η:AC I\eta : A \to C^I the right homotopy in question. We start by considering the homotopy fiber product

    D:=B× CC I× CB W B W tW C I d 0 C d 1 B tW C, \array{ D := B \times_C C^I \times_C B &\to&&\stackrel{\in W}{\to}& B \\ \downarrow^{\mathrlap{\in W}} &&&& \downarrow^{\mathrlap{t \in W}} \\ && C^I &\stackrel{d_0}{\to}& C \\ \downarrow && \downarrow^{d_1} \\ B &\stackrel{t \in W}{\to}& C } \,,

    where the long morphisms are weak equivalences by the lemma on morphisms out of homotopy fiber products.

Then consider the two universal morphisms

(f,η,g):AB× CC I× CB (f,\eta,g) : A \to B \times_C C^I \times_C B


(Id,σt,Id):BWB× CC I× CB (Id, \sigma \circ t, Id) : B \stackrel{\in W}{\to} B \times_C C^I \times_C B

into that. It follows by 2-out-of-3 that the latter is a weak equivalence. Factoring this using the factorization lemma produces hence

BWDWFD. B \stackrel{\in W}{\to} D' \stackrel{\in W \cap F}{\to} D \,.

We know moreover that the product map DFB×BD \stackrel{\in F}{\to} B \times B is a fibration, as we can rewrite the homotopy limit as the pullback

D C I F B×B f×g C×C. \array{ D &\to& C^I \\ \downarrow && \downarrow^{\mathrlap{\in F}} \\ B \times B &\stackrel{f \times g}{\to}& C \times C } \,.

It follows that the composite DDB×BD' \to D \to B \times B is a fibration and hence DD' a path space object for BB.

Finally, by setting A=A× DDA' = A \times_D D' we obtaine the desired right homotopy ftgtf \circ t' \simeq g \circ t'.

A D t A D C I f×g B×B t×t C×C. \array{ A' &\to& D' \\ \downarrow^{\mathrlap{t'}} && \downarrow \\ A &\to & D &\to & C^I \\ & {}_{f \times g}\searrow & \downarrow && \downarrow \\ && B \times B &\stackrel{t \times t}{\to}& C \times C } \,.

The relation “f,gC(A,B)f, g \in C(A,B) are homotopic”, fgf \sim g, is an equivalence relation on C(A,B)C(A,B).


The nontrivial part is to show transitivity. This now follows with the above lemma about homtopy commutative composition of pullback diagrams and then using the “piecing together of path objects” used above to show that right homotopy is an equivalence relation.


For CC a category of fibrant objects the category πC\pi C is defined to be the category

  • with the same objects as CC;

  • with hom-sets the set of equivalence classes

    πC(A,B):=C(A,B)/ \pi C(A,B) := C(A,B)/_\sim

    under the above equivalence relation.

  • Composition in πC\pi C is given by composition of representatives in CC.


The obvious functor

CπC C \to \pi C

is the identity on objects and the projection to equivalence classes on hom-set.

Let πWMor(πC)\pi W \subset Mor(\pi C) be the image of the weak equivalences of CC in πC\pi C under this functor, and πF\pi F the image of the fibrations.


The weak equivalences in πC\pi C form a left multiplicative system.


This is now a direct consequence of the above lemma on homotopy-commutative completions of diagrams.

The homotopy category

We discuss now that the structure of a category of fibrant objects on a homotopical category CC induces

  • a related category πC\pi C

  • with a morphism CπCC \to \pi C

    • that is the identity on objects,

    • and induces on πC\pi C a notion of weak equivalences

      πWMor(πC) \pi W \subset Mor(\pi C)

      and fibrations

      πFMor(πC) \pi F \subset Mor(\pi C)
  • such that

This implies the following convenient construction of the homotopy category of CC:


For CC a category of fibrant objects, its homotopy category is (equivalent to) the category Ho CHo_C with

  • the same objects as CC;

  • the hom-set Ho C(A,B)Ho_C(A,B) for all A,BObj(C)A, B \in Obj(C) given naturally by

    Ho C(A,B) colim A^wπWAπC(A^,B) =colim A^fπWFAπC(A^,B). \begin{aligned} Ho_C(A,B) & \simeq colim_{\hat A \stackrel{w\in \pi W}{\to} A} \pi C (\hat A,B) \\ & = colim_{\hat A \stackrel{f\in \pi W\cap F}{\to} A} \pi C (\hat A,B) \end{aligned} \,.

Here the colimit is, as described at multiplicative system, over the opposite category of the category πW A\pi W_A or (πFπW) A(\pi F\cap \pi W)_A whose objects are weak equivalences A^wπWA\hat A \stackrel{w \in \pi W}{\to} A or acyclic fibrations A^fπWFA\hat A \stackrel{f \in \pi W\cap F}{\to} A in πC\pi C, and whose morphisms are commuting triangles

A^ h A^ A \array{ \hat A &&\stackrel{h}{\to}&& \hat A' \\ & \searrow && \swarrow \\ && A }

in πC\pi C (i.e. for arbitrary hh).

So more in detail the above colimit is over the functor

πC(,B) A:(πW A) op(πC) opπC(,B)Set, \pi C(-, B)_A : (\pi W_A)^{op} \to (\pi C)^{op} \stackrel{\pi C(-, B)}{\to} Set \,,

where the first functor is the obvious forgetful functor.


It is again the factorization lemma above (and using 2-out-of-3 that implies that inverting just the acyclic fibrations in CC is already equivalent to inverting all weak equivalences. This means that the above theorem remains valid if the weak equivalences t:AAt : A' \to A are replaced by acyclic fibrations:

every cocycle Y g A f X\array{ Y &\stackrel{g}{\to}& A \\ {}^\simeq \downarrow^{\mathrlap{f}} \\ X }

out of a weak equivalence is refines by a cocycle out of an acyclic fibrantion, namely

E fX Y g A FW f X. \array{ \mathbf{E}_f X &\stackrel{\simeq}{\to}& Y &\stackrel{g}{\to}& A \\ &{}_{\in F \cap W}\searrow& {}^\simeq \downarrow^{\mathrlap{f}} \\ && X } \,.

Using acyclic fibrations has the advantage that these are preserved under pullback. This allows to consistently compose spans whose left leg is an acyclic fibration by pullback. See also the discussion at anafunctor.

A discussion of this point of using weak equivalences versus acyclic fibrations in the construction of the homotopy category is also in Jardine: Cocycle categories.

We now provide the missing definitions and then the proof of this theorem.


The homotopy categories of CC and πC\pi C coincide:

Ho CHo πC. Ho_C \simeq Ho_{\pi C} \,.

By one of the lemmas above, the morphisms d i:B IBd_i : B^I \to B are weak equivalences and become isomorphisms in Ho CHo_C. The section σ:BB I\sigma : B \to B^I then becomes an inverse for both of them, hence the images of d 0d_0 and d 1d_1 in Ho CHo_C coincide. Therefore the above diagram says that homotopic morphisms in CC become equal in Ho CHo_C.

But this means that the localization morphism

Q C:CHo C Q_C : C \to Ho_C

factors through πC\pi C as

Q C:CπCQ πCHo C Q_C : C \to \pi C \stackrel{Q_{\pi C}}{\to} Ho_C

where Q πCQ_{\pi C} sends weak equivalences in πC\pi C to isomorphisms in Ho CHo_C.

The universal property of QQ then implies the universal property for Q πCQ_{\pi C}

C πC A Q C Q πC Ho C. \array{ C &\to& \pi C &\to & A \\ \downarrow^{\mathrlap{Q_C}} & \swarrow^{Q_{\pi C}} && \swarrow \\ Ho_C } \,.

The above theorem on the description of Ho CHo_C now follows from the general formula for localization at a left multiplicative system of weak equivalences.

Pointed category of fibrant objects

If the category CC of fibrant objects has an initial object which coincides with the terminal object ee, i.e. a zero object, then CC is a pointed category. In this case we have the following additional concepts and structures.


For p:YXp : Y \to X a fibration, the pullback FF in

F i Y e X \array{ F &\stackrel{i}{\to}& Y \\ \downarrow && \downarrow \\ e &\to& X }

is the fibre of pp and ii is the fibre inclusion. (This is the kernel of the morphism ff of pointed objects)

Fibration Sequences

(See also fibration sequence)

For BB any object and B IB^I any of its path objects, the fiber of B Id 0×d 1B×BB^I \stackrel{d_0 \times d_1}{\to} B \times B is the loop object Ω (I)B\Omega^{(I)} B of BB with respect to the chosen path object. This construction becomes independent up to canonical isomorphism of the chosen path space after mapping to the homotopy category and hence there is a functor

Ω:Ho CHo C \Omega : Ho_C \to Ho_C

which sends any object BB of CC to its canonical loop object ΩB\Omega B.

Any loop object ΩB\Omega B becomes a group object in Ho CHo_C, i.e. a group internal to Ho CHo_C in a natural way.

Derived hom-spaces

There is an explicit simplicial construction of the derived hom spaces for a homotopical category that is equipped with the structure of a category of fibrant objects. This is described in (Cisinksi 10) and (Nikolaus-Schreiber-Stevenson 12, section 3.6.2).


For 𝒞\mathcal{C} a category of fibrant objects, write for any X,AObj(𝒞)X, A \in Obj(\mathcal{C})

Cocycle(X,A),wCocycle(X,A)Cat Cocycle(X,A) , wCocycle(X,A) \in Cat

for the categories (“categories of cocycles on XX with coefficients in AA”) whose objects are correspondences

XX^A X \stackrel{\simeq}{\longleftarrow} \hat X \stackrel{}{\longrightarrow} A

with the left leg an acyclic fibration (for Cocycle(X,A)Cocycle(X,A)) or just a weak equivalence (for wCocycle(X,A)wCocycle(X,A)); and whose morphisms are morphisms of spans

X^ X A X^ \array{ && \hat X \\ & \swarrow && \searrow \\ X && \downarrow && A \\ & \nwarrow && \nearrow \\ && \hat X' }

Write L we H𝒞L^H_{we} \mathcal{C} for the simplicial localization of the category of fibrant objects 𝒞\mathcal{C} at its weak equivalences (hence essentially the (infinity,1)-category that it presents). Then for all objects X,AObj(𝒞)X,A \in Obj(\mathcal{C}) the canonical maps

NCocycle(X,A)NwCocycle(X,A)L we H𝒞(X,A) N Cocycle(X,A) \to N wCocycle(X,A) \to L^H_{we} \mathcal{C}(X,A)

of simplicial sets (on the left the nerves of the cocycle categories of def. , on the right the derived hom space given by the simplicial localization) are weak homotopy equivalences.

In other words, NCocycle(X,A) wheNwCocycle(X,A)N Cocycle(X,A) \simeq_{whe} N wCocycle(X,A) is a model for the correct derived hom space.

From this it follows for instance that


The homotopy fiber products in 𝒞\mathcal{C} as defined in def. present indeed the correct (infinity,1)-limits.


Observe that for each object XX the 2-functor NCocycle(X,):𝒞sSetN Cocycle(X,-) \colon \mathcal{C} \to sSet of def. sends fibrations to Kan fibrations of simplicial sets (the horn-filling condition comes down to factoring maps through the given fibration, which is possible by pullback along the fibration). Moreover, it is evident that NCocycle(X,)N Cocycle(X,-) preserves ordinary pullbacks. This means that NCocycle(X,)N Cocycle(X,-) takes pullbacks along a fibration in 𝒞\mathcal{C} to pullbacks in sSet one of whose maps is a Kan fibration. Since the standard model structure on simplicial sets sSet QuillensSet_{Quillen} is a right proper model category, this means that these are homotopy pullbacks (as discussed there) in sSet QuillensSet_{Quillen}. Finally by prop. this means that the derived hom-space functor Hom(X,)\mathbb{R}Hom(X,-) sends pullbacks along fibrations to homotopy pullbacks of the correct derived hom-spaces. This means (as discussed for instance at homotopy Kan extension) that the original pullbacks in 𝒞\mathcal{C} are the correct homotopy pullbacks.

Application in cohomology theory

When the catgegory of fibrant objects is that of locally Kan simplicial sheaves, the hom-sets of its homotopy category compute generalized notions of cohomology.

At abelian sheaf cohomology is a detailed discussion of how the ordinary notion of sheaf cohomology arises as a special case of that.


The notion of category of fibrant objects was introduced and the above results obtained in

for application to homotopical cohomology theory.

A review is in section I.9 of

There is a description and discussion of this theory and its dual (using cofibrant objects) in

Discussion of embeddings of categories of fibrant objects into model categories is in

Also discussion of the derived hom spaces in categories of fibrant objects is in that article, as well as in section 6.3.2 of

and also in

Usage of categories of fibrant objects for the homotopical structure on C*-algebras is in :

Categories of fibrant objects form a convenient setting for the study of homotopy type theory:

  • Jeremy Avigad, Chris Kapulkin, Peter LeFanu Lumsdaine, Homotopy limits in type theory 1304.0680

It is shown in the following paper that categories of fibrant objects are themselves fibrant in the model structure on categories with weak equivalences:

  • Lennart Meier, Fibration Categories are Fibrant Relative Categories, arxiv

Last revised on February 23, 2024 at 23:16:08. See the history of this page for a list of all contributions to it.