A discourse analysis of the English tense-aspect-modality system

This study presents three empirically grounded discourse analyses of the English tense-aspect-modality markers that encode past habituality (used to vs. would), past events (the present perfect vs. the preterit), and futurity (be going to vs. will). The discourse functions of these forms are analyzed in spoken narratives and spontaneous conversations with special reference to the discourse-pragmatic and semantic features of these forms.; The point of departure for the study is the "Frame-Elaboration Hypothesis", which draws upon my preliminary observations that used to, the present perfect, and be going to tend to set up a frame for the following discourse, thus often marking a point where an episode is initiated. In contrast, their competing forms, i.e., would, the past tense, and will, are used for subsequent elaboration within the frame established by the previously mentioned item in each pair. While some differences are noted regarding the ways in which each form behaves as opposed to its competing form in local contexts, these findings strongly support the Frame-Elaboration Hypothesis in terms of how speakers employ the competing forms in organizing discourse and managing interaction.; The discourse-pragmatic analysis of the forms in question constitutes a basis on which the semantic properties of these forms are explicated. The semantic accounts, which are pragmatically grounded, focus on the speaker's communicative motivation for initiating an episode by way of anchoring a past or future event to the here and now. The notion of "current relevance" is proposed as the semantic feature that directly bears upon the semantic opposition between the present perfect and the preterit, and the notion of "present orientation" is introduced to distinguish be going to from will in terms of marking futurity. As for used to and would, which are not treated at the same semantic level, I note that "past habituality" with "remoteness" is inherent to used to, while past habituality is a derived message in the case of would.; These findings suggest that the major differences between the competing forms, which cut through the three pairs of TAM markers, can be captured in a unified manner when we consider the discourse-pragmatic-interactional features as well as their semantic features.