In China, Japan, and Taiwan nowadays, there exists an institutionalized national language, discursively constructed as a homogeneous entity. This framing makes it challenging to critically examine monolingualism and its impact on linguistic minorities, hybrids, and alternative language practices. To excavate the alternatives that were hardly examined in monolingualist historical narrative, in this dissertation, I focus on the hybrid language practices of the 1920s and 1930s. While each of the four chapters explores distinct subjects, they all share a common thread—the utilization of hybrid languages. These practices encompass languages used on screen, such as bilingual intertitles (Chapter 1), and offscreen, including benshi’s oral performances (Chapter 2). Additionally, they were also embodied in modernist literature either written in or inspired by the hybrid practices of languages (Chapter 3) and (mis-)translation (Chapter 4). During the 1920s and 1930s, hybrid or heterolingual practices were widespread. However, due to shrinking creative spaces, increasing nationalist sentiments, and growing centralized government control, these practices were gradually suppressed, albeit they still survive in elusive and yet variegated ways. These practices not only creatively responded to geopolitics or local politics by reflecting on the legacies of previous language movements, but they also offered new ways to imagine community. They serve as a microcosm of multi-layered power dynamics, spanning geopolitical, transnational, domestic, historical, and ideological realms.