One way to see that stabilization of a Heegaard diagram produces the same manifold comes from Morse theory. Consider \(\Sigma_g\) as the level set of a self-indexing Morse function \(f\). Suppose that we wanted to modify our Morse function to \(\tilde f\) by adding in a pair of critical points \(p, q\) so that \(\ind(p)=1\) and \(\ind(q)=2\). We imagine that the critical points would appear on opposite sides of \(\Sigma_g\), and be connected by a single flow line. Furthermore, \(\tilde \Sigma_{g+1}=\tilde f(1.5)\), the new level set, would be of genus \(g+1\). By applying surgery along either the attaching circles \(W^\downarrow(p)\cap \tilde \Sigma_{g+1}\) or \(W^\uparrow(q)\cap \tilde \Sigma_{g+1}\), we obtain \(\Sigma_g\). See figure 0.0.1.
figure 0.0.1:From the perspective of Morse theory, stabilization comes from the creation of a pair of cancelling critical points.