This paper proposes a punctuation analysis framework centered on the notion of sentence fields and the flow of tension, offering a re-examination of the roles of the period, ellipsis, and em dash at sentence-final positions. Rather than treating punctuation as an addition of tone, a cue for pause, or a purely conventional writing device, this paper argues that punctuation actively participates in the formation of boundary conditions within a sentence field, and systematically alters how tension flows within that field. From this perspective, the differences among the period, ellipsis, and em dash should not be described merely in terms of “whether a sentence ends” or “degrees of emphasis.” Instead, they are better understood as three distinct topologies of tension flow. The period produces a sustained closed-loop circulation of tension within a bounded field. The ellipsis allows tension to extend along its original trajectory, maintaining an open and unconverged state. The em dash, by contrast, forces a connection between otherwise separable sentence fields, enabling tension to cross the original boundary and enter a shared flow structure. The distinction among these three does not lie in the presence or absence of tension, but in how tension is closed, extended within a field, or connected across fields. On this basis, the paper further argues that although punctuation is often omitted in actual writing, its structural function does not thereby disappear. The period and ellipsis primarily operate on the boundary conditions of a single sentence field; even when omitted, readers can often infer the general direction of tension from context. The em dash differs in that it does not merely adjust a single field, but enforces a connection between otherwise separable sentence fields, causing them to be read as a unified field. For this reason, when the em dash is omitted, readers are more likely to perceive a structural irregularity, rather than a simple shift in tone. To articulate this framework, the paper characterizes the period, ellipsis, and em dash respectively as three fundamental modes: closed-loop, extension, and connection. Through minimal contrasts and schematic descriptions, it demonstrates how different punctuation marks alter boundary conditions, tension flow, and field connectivity. It is thus argued that these punctuation marks should not be regarded as secondary symbols external to content, but as boundary operations that actively participate in the formation of sentence field structures.
Rinelle Chen (Mon,) studied this question.