Many early Chinese writings feature a scene of speech. In more literary writings as in more technical ones, writing often puts in the mouth of a speaker what the text tries to get across. Frequently, the speaker’s spoken words, comprising a text either in full or in part, also occupy the center stage of writing’s attempt at narrating a story, arguing a point, imparting a skill, or simply sorting things out. The list goes on. This lasting, shared, and versatile appeal to, and of, the spoken word in early Chinese writings is the topic of this dissertation. It argues that speech representation is first and foremost a narrative device serving a text’s rhetorical aims. Writing does not represent speech as a mere technology of transcription; rather, writing creates a world-in-itself with and through the representation of speech, a highly elevated medium of expression since early on. In the process, this dissertation also shows, there developed a set of representational structures and narrative techniques with which speech representation was approached. The former supply writing with the forms, shapes, and tropes that underlie the representation of speech; the latter the means to make creative but circumscribed variations. Together, they cement the role of the spoken word as a narrative construct that different writings can use and adapt for their own rhetorical purposes. In highlighting these aspects of the spoken word, this dissertation offers an account of a core but underemphasized attribute of early Chinese textuality, models a new methodological approach, and proposes a range of interpretations of texts old and new intended to reevaluate their meaning and significance. Its goal, ultimately, is to promote a new way of reading.