Proving that induced homomorphism is an isomorphism

583 Views Asked by At

Let $A \subset X$ and let $j: A \to X$ be the inclusion map. Let $f: X \to A$ be a continuous map. We suppose there is a homotopy between $j \circ f$ and the identity map on $X$. We are to show that if $f$ is a retraction, then $j_*$ (induced homomorphism) is an isomorphism.

There seems to be a fundamental misunderstanding on my part, in this scenario, I think of fixed point(s) $a, j(a) = a$ and consider $j$ as a map $$j: (A, a) \to (X, a)$$ and think of the induced homomorphism as $$j_*: \pi_1(A, a) \to \pi_1(X,a) \\ [g] \mapsto [j \circ g]$$ while the solutions I found online consider it as a map $$j_*: [A, X] \to [X, X]$$ where $[X, Y]$ denotes the set of homotopy classes from $X$ to $Y$. While such a view does, to a certain degree make sense in this setting I must admit I fail to see why its allowed. To me it seems that we are no longer considering members of the equivalence class, but rather the equivalence classes themselves.

1

There are 1 best solutions below

2
On BEST ANSWER

Maybe this solution clears it up for you:

We have that $f:X \to A$ is a retraction. So that then the restriction $f|_{A}$ is the identity map on $A$($\subset X$).

As $j: A \to X$, then the composition $f \circ j$ is the identity map on $A$. It follows that this composition induces a homomorphism $f_{*}\circ j_{*}$ which is the identity map on $\pi_{1}(A,a)$.

In particular, for any loops $[k],[l] \in \pi_{1}(A,a)$ such that $j_{*}[k]=j_{*}[l]$, then we have

$(f_{*}\circ j_{*})[k]=(f_{*} \circ j_{*})[l]$ implies $[k]=[l]$, as $f_{*}\circ j_{*}$ is the identity map on $\pi_{1}(A,a).$ It follows at once that $j_{*}$ is injective.

Note that, it is not necessarily true that $j_{*}$ is a surjection. Thusly, it's not necessarily an isomorphism.