Previous studies directly found that the language control network in humans has neuroplasticity. Nevertheless, its stability of it has not been investigated directly. It is poorly understood whether and how these two seemingly opposing features are implemented within the same network. In the present study, we first directly measured the stability and neuroplasticity of the language control network by resting-state functional connectome. Then we explored morphological and topological coupling patterns and the predictive abilities of two features. We employed a cohort of Chinese first-year college students majoring in English who underwent second language (L2) acquisition courses at university and another cohort of Chinese students who did not undergo the same course. Two resting-state fMRI scans were acquired approximately one year apart. We found that the entire language control network is stable rather than plastic. And the morphological coupling pattern was not modular. The dorsal anterior cingulate cortex (dACC), the left inferior frontal gyrus (IFG), and the right Caudate showed a special role in the topological coupling pattern. Moreover, richer second language learning suggested lower modular coupling and different topological coupling patterns. Interestingly, the FC part for stability could predict the improvement of L2 proficiency. These findings provide preliminary evidence about how stability and plasticity coexist in the same neural network and deepen our understanding of the development of the language control network in long-term second language learning.