Information, expectation, and processing: On classifying poetic texts

Abstract

This paper is concerned with the question of text type classification on the basis of criteria of information and expectations, as well as the psychological strategies for comprehending and processing texts. The notion ‘text type’ is defined as: a set of stored instructions for gathering and processing the discernible, multi-level features of texts. The set is activated or ‘addressed’ when an actual text matches the typical parameters or the ‘frames’ of expectations concerning a text type, following a preliminary scanning. Two sample texts — a poem and a science textbook — are analysed by means of expectation flow diagrams that show how textual patterns are established on various ‘levels’ such as sound, grammar and lexicon. Even when the discernible features of two texts of opposed types are strikingly similar (as is in fact the case with the samples here), different processing priorities are postulated. A further issue is the next-type-specific instruction for truth-value assignment, which the authors studied have apparently incorporated into their planning of author-reader interaction.