+-- {: .rightHandSide} +-- {: .toc .clickDown tabindex="0"} ###Context### #### Homological algebra +--{: .hide} [[!include homological algebra - contents]] =-- =-- =-- #Contents# * table of contents {:toc} ## Definition {#Definition} ### In an abelian category {#InAbelianCategory} Let $\mathcal{A}$ be an [[abelian category]]. +-- {: .num_defn #SplitnessInAbelianCategory} ###### Definition A [[short exact sequence]] $0\to A \stackrel{i}{\to} B \stackrel{p}{\to} C\to 0$ in $\mathcal{A}$ is called **split** if either of the following equivalent conditions hold 1. There exists a [[section]] of $p$, hence a morphism $s \colon C\to B$ such that $p \circ s = id_C$. 1. There exists a [[retract]] of $i$, hence a morphism $r \colon B\to A$ such that $r \circ i = id_A$. 1. There exists an [[isomorphism]] of sequences with the sequence $$ 0\to A\to A\oplus C\to C\to 0 $$ given by the [[direct sum]] and its canonical injection/projection morphisms. =-- +-- {: .num_lemma #SplittingLemma} ###### Lemma **(splitting lemma)** The three conditions in def. \ref{SplitnessInAbelianCategory} are indeed [[equivalence|equivalent]]. =-- (e.g. [Hatcher (2002)](#Hatcher02), p. 147) +-- {: .proof} ###### Proof It is clear that the third condition implies the first two: take the section/retract to be given by the canonical injection/projection maps that come with a [[direct sum]]. Conversely, suppose we have a retract $r \colon B \to A$ of $i \colon A \to B$. Write $P \colon B \stackrel{r}{\to} A \stackrel{i}{\to} B$ for the corresponding [[idempotent]]. Then every element $b \in B$ can be decomposed as $b = (b - P(b)) + P(b)$ hence with $b - P(b) \in ker(r)$ and $P(b) \in im(i)$. Moreover this decomposition is unique since if $b = i(a)$ while at the same time $r(b) = 0$ then $0 = r(i(a)) = a$. This shows that $B \simeq im(i) \oplus ker(r)$ is a [[direct sum]] and that $i \colon A \to B$ is the canonical inclusion of $im(i)$. By exactness it then follows that $ker(r) \simeq im(p)$ and hence that $B \simeq A \oplus C$ with the canonical inclusion and projection. The implication that the second condition also implies the third is formally dual to this argument. =-- ### In a semi-abelian category There is a nonabelian analog of split exact sequences in [[semiabelian categories]]. See there. ## Properties ### Relation to chain homotopy {#RelationToChainHomotopy} +-- {: .num_prop} ###### Proposition A [[long exact sequence]] $C_\bullet$ is _split exact_ precisely if the [[weak homotopy equivalence]] from the 0-chain complex, namely the [[quasi-isomorphism]] $0 \to C_\bullet$ is actually a [[chain homotopy]] [[homotopy equivalence|equivalence]], in that the [[identity]] on $C_\bullet$ has a [[null homotopy]]. =-- ### Of free modules and vector spaces {#OfVectorSpaces} Assuming the [[axiom of choice]]: +-- {: .num_prop} ###### Proposition Every [[exact sequence]] of [[free abelian groups]] is split. =-- +-- {: .num_prop} ###### Proposition Every exact sequence of [[free modules]] which is [[bounded-below chain complex|bounded below]] is split. =-- Let $k$ be a [[field]] and denote by $\mathcal{A} \coloneqq k$[[Vect]] the [[category]] of [[vector spaces]] over $k$. +-- {: .num_cor #SESOfVectorSpacesSplits} ###### Corollary Every [[short exact sequence]] of vector spaces is split. =-- (Essentially by the [[basis theorem]], for exposition see for instance [here](https://unapologetic.wordpress.com/2008/06/26/exact-sequences-split).) ### Involving injective/projective objects {#InvolvingInjectiveObjects} +-- {: .num_lemma } ###### Lemma If in a [[short exact sequence]] $0 \to A \to B \to C \to 0$ in an [[abelian category]] the first object $A$ is an [[injective object]] or the last object is a [[projective object]] then the sequence is split exact. =-- +-- {: .proof} ###### Proof Consider the first case. The other is formally dual. By the properties of a [[short exact sequence]] the morphism $A \to B$ here is a [[monomorphism]]. By definition of [[injective object]], if $A$ is injective then it has the [[right lifting property]] against [[monomorphisms]] and so there is a morphism $q : B \to A$ that makes the following [[diagram]] [[commuting diagram|commute]]: $$ \array{ A &\stackrel{id_A}{\to}& A \\ \downarrow & \nearrow_{q} \\ B } \,. $$ Hence $q$ is a [[retract]] as in def. \ref{SplitnessInAbelianCategory}. =-- ## References For instance * [[Charles Weibel]], Section 1.4 of: _[[An Introduction to Homological Algebra]]_ (1994) * {#Hatcher02} [[Allen Hatcher]], pp. 147 of: *Algebraic Topology*, Cambridge University Press (2002) [[ISBN:9780521795401](https://www.cambridge.org/gb/academic/subjects/mathematics/geometry-and-topology/algebraic-topology-1?format=PB&isbn=9780521795401), [webpage](https://pi.math.cornell.edu/~hatcher/AT/ATpage.html)] [[!redirects split sequence]] [[!redirects split exact sequences]] [[!redirects split sequences]] [[!redirects split short exact sequence]] [[!redirects split short exact sequences]] [[!redirects splitting lemma]]