From homomorphism to covering. Given \(\rho \colon \pi_1(X, x_0) \to G\text{,}\) let \(\pi_1(X, x_0)\) act on \(\tilde{X} \times G\) by
\begin{equation*}
[\sigma] \cdot (z, g) = \bigl([\sigma] \cdot z,\; g \cdot \rho([\sigma])^{-1}\bigr),
\end{equation*}
where
\([\sigma] \cdot z\) is the action from
Theorem 4.11. Set
\(Y_\rho = (\tilde{X} \times G)/\pi_1(X, x_0)\text{.}\) The map
\(P_\rho \colon Y_\rho \to X\) is a
\(G\)-covering, where
\(G\) acts on
\(Y_\rho\) by
\(h \cdot \langle z, g\rangle = \langle z, hg \rangle\text{.}\)
From covering to homomorphism. Given a \(G\)-covering \(p \colon Y \to X\) with basepoint \(y_0\) over \(x_0\text{,}\) define \(\rho \colon \pi_1(X, x_0) \to G\) by
\begin{equation*}
\rho([\sigma]) = \text{endpoint of the lift of }\sigma\text{ starting at }y_0,
\end{equation*}
viewing the endpoint as an element of the \(G\)-fibre over \(x_0\text{.}\) These two constructions are inverse to each other.