Accentuation and Interpretation

Optimal accentuation of a sentence involves accentuating a minimal set of words which in a given context suffices for understanding the entire sentence. We propose a model of the interpretation of incomplete or not entirely recognized utterances. Using this model, we determine which constituents of an utterance have to be accentuated given a certain context. 1. Optimal Accentuation Our basic assumption is that accentuation helps to ensure the listener’s recognition of certain words. These words are sufficent for interpreting the entire utterance in a given context. The recognition of other words is not necessary for understanding. We propose a model for the interpretation of not entirely recognized utterances, like acoustically disturbed utterances, and of intrinsically incomplete utterances, like constituent-answers and telegrams. Speakers want listeners to understand their utterances. They utter their sentences in a way that leads the listeners to the intended interpretations. Accentuation is driven by the listeners’ interpretative needs. It is one means of following Grice’s maxim of manner ([2]) “Be perspicuous!”. For the correct interpretation of an utterance, the listener has to be aware of its context. Knowledge of the current issue or question under discussion (qud) (cf. [1]) is needed in order to derive the exhaustive meaning of an utterance. Consider example 1.1: