The Representation of the Temporal Properties of Linguistic Utterances through a Mathematical Model

Abstract

In this article we use a mathematical model to encode the temporal properties of linguistic utterances across languages by means of mathematical objects—points, lines, segments, vectors and versors—and the relations established among them in a four-dimensional space. Such temporal properties are encoded through threedifferent systems: tense—past, present and future—which locates the utterance on a temporal line, aspect—perfectivity and progressivity—which sets the viewpoint of the speaker, and Aktionsart, which refers to the structural temporal properties of the utterance such as telicity—whether the event has an endpoint or not—dynamicity—whether a change is conveyed or not—and duration. This model aims to be language independent in order to allow for the codification of the temporal properties of utterances in any language, thus rendering it appropriate to be used as an interlingua in Natural Language Processing (NLP) applications. This wouldsignificantly improve the comprehension of natural language in search engines and automatic translation systems, to name two examples. Hence, our ultimate goal is for this model to achieve computational adequacy.