In this essay, I reread Naoki Sakai’s Voices of the Past (1992) in order to examine its theoretical implications and to critically engage with the debate it has evoked. I focus on his analysis of how a new regime of translation gave rise to the discourse on a national language in eighteenth-century Japan. I first explain the hybrid nature of kanbun kundoku, a technique of making classical Chinese read like Japanese. I then discuss the Japanese Confucian scholar Ogyu Sorai, showing how his new phonetic method of reading and translating kanbun was implicated in eliminating kundoku and drawing the very boundary between “Japanese” and “Chinese.” I also examine the debate concerning Sakai’s interpretation of Sorai as a phonocentric thinker. Although some scholars insist that Sorai emphasized the visual aspect of Chinese texts and thereby abandoned the phonetic method, I argue that Sorai’s approach to writing was still modeled after speech and the act of enunciation. Finally, I explain his practice of translation in terms of what Sakai calls the “schema of cofiguration,” a regime of constructing linguistic and cultural binary oppositions, such as Japanese and Chinese, the “West” and the “non-West,” and so forth.