Understanding bilingual language processing and control requires approaches that move beyond correlational and univariate analyses. This talk introduces how Representational Similarity Analysis (RSA) and Transcranial Magnetic Stimulation (TMS) offer new insights into the bilingual brain. Drawing on findings from our experimental studies, I will show how RSA reveals representational structure across activation, connectivity, and behavior—highlighting hierarchical language representations, neural reconfiguration across domains, and differences between production and comprehension. In contrast, TMS provides converging causal evidence. Offline stimulation of the right inferior frontal gyrus (rIFG) disrupts both behavioral efficiency and neural activity during language switching, demonstrating its necessity for inhibitory control. Online TMS further uncovers dynamic, time-sensitive interactions between the language control and representation systems. Together, these studies illustrate how integrating multivariate representational analysis with causal neuromodulation deepens our understanding of bilingual language representation and control, and highlights future directions involving machine learning and brain–AI representational alignment.